Citation
Investigation of double bevel full penetration groove welds for the repair of historic structural wrought iron

Material Information

Title:
Investigation of double bevel full penetration groove welds for the repair of historic structural wrought iron
Creator:
Watters, Joel D. ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (113 pages). : ;

Subjects

Subjects / Keywords:
Wrought-iron ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Structural wrought iron was used extensively in the second half of the 19th century to construct bridges and buildings throughout the United States. Many of these wrought iron structures are still in use today, and are in need of repair and renovation. While many methods re used for the repair of these structures, there is no published, pre-qualified welding procedure using modern welding technology for the repair of historic structural wrought iron. This thesis investigates the ability of a specific welding procedure, using modern welding technology to repair structural wrought iron samples which were removed from a historic bridge. Additionally, the physical properties of this historic structural wrought iron were tested and documented, in an effort to increase the data available to engineers who are tasked to work with this unique structural material.
Thesis:
Thesis (M.S.)--University of Colorado Denver. Civil engineering
Bibliography:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Joel D. Watters.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
892630915 ( OCLC )
ocn892630915

Downloads

This item has the following downloads:


Full Text
INVESTIGATION OF DOUBLE BEVEL FULL PENETRATION GROOVE WELDS
FOR THE REPAIR OF HISTORIC STRUCTURAL WROUGHT IRON
by
JOEL D. WATTERS
B.S., University of Southern California, 2007
A thesis submitted to the
Faculty of the Graduate School
University of Colorado
in partial fulfillment of the requirements for the degree of
Masters of Science
Civil Engineering
2013


This thesis for the Master of Science Degree
by
Joel D. Watters
has been approved for the
Department of Civil Engineering
by
Frederick Rutz, Chair
Kevin Rens
Cheng Yu Li
November 12, 2013


Watters, Joel D. (M.S., Civil Engineering)
Investigation of Double Bevel Full Penetration Groove Welds for the Repair of
Historic Structural Wrought Iron
Thesis directed by Assistant Professor Frederick R. Rutz
ABSTRACT
Structural wrought iron was used extensively in the second half of the 19th
Century to construct bridges and buildings throughout the United States. Many of
these wrought iron structures are still in use today, and are in need of repair and
renovation. While many methods are used for the repair of these structures,
there is no published, pre-qualified welding procedure using modern welding
technology for the repair of historic structural wrought iron. This thesis
investigates the ability of a specific welding procedure, using modern welding
technology, to repair structural wrought iron samples which were removed from a
historic bridge. Additionally, the physical properties of this historic structural
wrought iron were tested and documented, in an effort to increase the data
available to engineers who are tasked to work with this unique structural metal.
The form and content of this abstract are approved. I recommend its
publication
Approved: Frederick Rutz


ACKNOWLEDGEMENTS
This thesis and the research conducted for its completion were made
possible by the tireless dedication of Dr. Frederick R. Rutz. Through his efforts,
large amounts of historic structural wrought iron were made available for this
important research, and his passion and dedication have enabled continued
research in the hopes of developing suitable welding procedures for this unique
structural metal.
Thank you to the faculty and welding students of Western Colorado
Community College of Grand Junction, CO for their support and investment of
time into this research. Their excellent work and efforts made the completion of
this thesis possible.
Thank you to Ryan Thomas of Emily Griffith Opportunity School in Denver,
CO for his invaluable training in the shielded metal arc welding process, and for
his knowledgable advice regarding the research conducted for this thesis.
Thank you to Tom Thuis and Jac Corless of the Civil Engineering lab, and
machine shop at University of Colorado Denver. They provided extensive
assistance in the preparation of wrought iron test specimens, and in the
operation of the testing equipment used.
IV


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION 1
Necessity of Research 1
Scope of Research 2
Goals for Research 3
II. HISTORY & LITERATURE REVIEW 5
History of Wrought Iron 5
Manufacture of Wrought Iron 5
Structural Applications of Wrought Iron 7
Properties of Wrought Iron 8
Minnesota Bridge #5721 15
Repair & Preservation of Structural Wrought Iron 17
Testing of Historic Structural Wrought Iron 18
Welding Repair of Structural Wrought Iron 19
III. WELDING BACKGROUND 25
Welding Methods 25
Shielded Metal Arc Welding 25
Gas Metal Arc Welding 29
Gas Tungsten Arc Welding 29
Standard Joint Types 30
Weld Quality, Flaws & Testing 32
Welding Procedure Qualification 35
v


IV. WROUGHT IRON TESTING TEST DESIGN & CONFIGURATION 37
Base Metal Tension Tests Design & Configuration 37
Test with 220k MTS 37
Test with 20k MTS 44
Base Metal Spark Test Design & Configuration 46
Welded Specimen Tension Tests Design & Configuration 46
Welded Specimen Bend Tests Design & Configuration 50
V. WROUGHT IRON TESTING RESULTS 56
Base Metal Testing Results 56
Spark Test Results 60
Welded Specimen Tension Tests Results 61
Welded Specimen Bend Tests Results 65
VI. DISCUSSION FROM TEST RESULTS 68
Base Metal Testing Discussion 68
Spark Test Discussion 71
Welded Specimen Tension Tests Discussion 72
Welded Specimen Bend Tests Discussion 74
VII. CONCLUSIONS & RECOMMENDATIONS 76
Conclusions from Research 76
Recommendations for Future Research 78
Welding Methods, Materials 78
Joint Types 79
Wrought Iron Material Investigation 79
VI


REFERENCES
81
APPENDIX
A. Tension Test Plots 82
B. Photographs Tension Tests 99
C. Photographs Bend Tests 102
VII


TABLES
TABLE
2.1 Typical Wrought Iron Chemical Analysis 14
2.2 Early 20th Century Steel Chemical Analysis 15
2.3 Results of Tensile Tests by Gordon & Knopf 19
2.4 Results of Tensile Tests by Chomsrimake 21
2.5 Results of Tensile Tests by Bowman & Piskorowski 23
3.1 Carbon Steel Covered Arc Welding Electrodes 28
3.2 Applicable Welds for Basic Joint Types 32
4.1 Weld Parameters for Welded Tension Tests 48
4.2 Weld Parameters for Welded Bend Tests 51
5.1 Base Metal Tension Test Results 57
5.2 Welded Specimen Tension Test Results 63
5.3 Welded Specimen Tension Test Results By Groove Angle 63
5.4 Welded Specimen Tension Test Results By Preheat Temperature 64
5.5 Welded Specimen Bend Test Results 66
viii


FIGURES
FIGURE
2.1 Wrought Iron Photomicrograph 9
2.2 Ruptured Wrought Iron Specimen 10
2.3 Wrought Iron Physical Properties 12
2.4 Minnesota Bridge 5721 Stillwater, MN 16
2.5 Single V-Groove Weld 20
3.1 Standard Joint Types 31
4.1 220k MTS Testing Frame 38
4.2 20k MTS Testing Frame 38
4.3 Design of Steel Clamp Bar for Tension Test 40
4.4 Design of Wrought Iron Tension Test Specimen 42
4.5 Base Metal Tension Test Setup 43
4.6 Wrought Iron Specimen for 20k MTS with Extensiometer 45
4.7 Eye-Bar M1 Segmentation Diagram 47
4.8 Double Bevel Full Penetration Groove Weld 49
4.9 Lincoln Electric Precision TIG 185 49
4.10 Eye-Bar M5 Segmentation Diagram 50
4.11 Preparation of Bend Test Specimen 52
4.12 Prepared Bend Test Specimens 53
4.13 Guided Bend Test Schematic 54
4.14 Watts W-50 Tensile & Bend Weld Tester 55
5.1 B-M1-1 Force vs. Displacement Plot 58
IX


5.2 B-M1-2: Stress vs. Strain Plot 59
5.3 Spark Test Photograph Wrought Iron from MN Bridge 5721 60
5.4 Spark Test Photograph Mild Steel 61
5.5 Welded Specimen Tension Test Photograph 1 64
5.6 Welded Specimen Tension Test Photograph 2 65
5.7 Welded Specimen Successful Root Bend Test 67
5.8 Welded Specimen Failed Face Bend Test 67
6.1 Visible Ductility Base Metal Tension Test 70
6.2 Spark Test Diagram 71
A.1 Specimen B-M3-2 Force vs. Displacement Plot 82
A.2 Specimen B-M1-1 Force vs. Displacement Plot 83
A.3 Specimen B-M1-2 Force vs. Displacement Plot 84
A.4 Specimen B-M3-3 Force vs. Displacement Plot 85
A.5 Specimen B-M3-4 Force vs. Displacement Plot 86
A.6 Specimen B-M1-1 Force vs. Displacement Plot 87
A.7 Specimen B-M1-2 Force vs. Displacement Plot 88
A.8 Specimen B-M1-3 Force vs. Displacement Plot 89
A.9 Specimen B-M1-4 Force vs. Displacement Plot 90
A.10 Specimen B-M1-5 Force vs. Displacement Plot 91
A.11 Specimen B-M1-6 Force vs. Displacement Plot 92
A.12 Specimen B-M1-7 Force vs. Displacement Plot 93
A.13 Specimen B-M1-8 Force vs. Displacement Plot 94
A.14 Specimen B-M1-9 Force vs. Displacement Plot 95
X


A.15 Specimen B-M5-10 Force vs. Displacement Plot 96
A.16 Specimen B-M5-11 Force vs. Displacement Plot 97
A.17 Specimen B-M5-12 Force vs. Displacement Plot 98
B.1 Tension Test with 220k MTS 99
B.2 Tension Test Specimen 99
B.3 Tension Test with 220k MTS 100
B.4 Tension Test with 20k MTS 101
C.1 Prepared Bend Test Specimens 102
C.2 Bend Test Operation 102
C.3 Failed Face Bend Test 103
C.4 Successful Root Bend Test 103
XI


ABBREVIATIONS
Avg. Average
AWS American Welding Society
CV Coefficient of Variation (Std. Dev. / Mean)
GMAW Gas Metal Arc Welding
GTAW Gas Tungsten Arc Welding
k kip, a unit of force, (1 k = 1000 lbs)
SMAW Shielded Metal Arc Welding
Std. Dev. Standard Deviation
WPS Welding Procedure Specification
WPQR Welding Procedure Qualification Record
XII


CHAPTER I
INTRODUCTION
Necessity of Research
In recent years the deteriorating state of infrastructure in the United States
has become increasingly clear. As roads, bridges and other infrastructure
components that were built over 100 years ago begin to require more and more
maintenance, it is vital that the means for repair and rehabilitation of existing
structures are in place. Due to limited budgets for infrastructure spending, it
cannot be expected that deteriorating infrastructure components can always be
replaced with new ones. Although much of the infrastructure in disrepair today is
built of concrete and steel, there are still a large number of structures, mostly
bridges, in the United States that incorporate structural wrought iron as a
significant part of their structural system. These bridges, most of which were built
over 100 years ago carry rail, vehicle and/or pedestrian traffic. Although little
construction has utilized structural wrought iron in the past century, it is important
for the engineering and construction community to have access to a broad range
of repair methods for the rehabilitation of existing wrought iron bridges.
Modern welding techniques have the potential to give engineers and
preservation contractors many ways to repair historic structural wrought iron.
Although welding has been performed on historic structural wrought iron, there is
a severe shortage of documented testing and approval of welding procedures for
this unique structural metal. Because most historic wrought iron structures are
1


bridges, their repair falls under the oversight of the local or state department of
transportation. For repairs that will affect the structural system of an existing
bridge, these governing bodies typically require that any welding performed
conforms to a pre-qualified welding procedure. The majority of pre-qualified
welding procedures published by the American Welding Society (AWS) apply to
steel and other modern structural metals, and approved repair methods using
modern welding techniques on historic structural wrought iron are unavailable to
engineers and contractors. For this reason, the research and development of
welding procedures for historic structural wrought iron is vital to the preservation
of existing infrastructure.
Scope of Research
Although the possibilities of welding historic wrought iron are very
extensive, the scope of research presented herein is limited due to time and
resources available within the Masters Thesis program. As discussed in more
detail in Chapter III, many welding methods may have potential use for the repair
of historic wrought iron. However, the research conducted for this thesis was
limited to a specific joint type and a specific welding rod. Due to the iterative
nature of this research, it was paramount that as many variables as possible
were kept constant, ideally only changing one variable at a time. For this project,
an E7018 welding rod was used, and the weld type was a double-bevel full
penetration groove weld on a butt joint (reference Chapter III). The parameters
that were iteratively adjusted included the pre-heat temperature of the joint, and
2


the groove angle of the weld joint. For a full description of the welding procedures
which were tested, refer to Chapter IV.
In addition to limiting the welding methods and joint types used, the tests
performed on the base metal and the welded specimens were limited to tension
tests, root bend tests, face bend tests, and spark tests. For more information
regarding the design and configuration, and the results of these tests, refer to
Chapters IV and V.
Goals for Research
The goal for this project is to develop a satisfactory welding procedure to
be used in the repair of historic structural wrought iron. Although simply stated, to
prove that a specific welding procedure will provide the required structural
integrity, it must be shown that the weld fully develops the strength of the base
metal. Additionally, it must be shown that the welding process does not
significantly reduce the ductility of the base metal. For an existing wrought iron
structural member subject to flexural stresses, reduction in ductility could lead to
premature, and possible brittle failure of the member. The tension tests described
in Chapter IV of this thesis aim to demonstrate that the weld performed will not
reduce the strength or ductility of the wrought iron specimen, and will not change
the failure mechanism of the specimen. The bend tests are designed to properly
qualify the welding procedure and to verify proper welding technique. In short,
the goal of the tests performed was to show that the full penetration weld of two
wrought iron pieces will allow the final welded specimen to behave the same or in
3


an improved manner when compared to the behavior of an un-welded specimen
of the same material.
4


CHAPTER II
HISTORY & LITERATURE REVIEW
History of Wrought Iron
Manufacture of Wrought Iron
Before steel became the primary structural metal for the construction
industry, wrought iron had, for the previous 50 years, been used to construct
buildings and bridges. Although all three of these products are made from iron,
and to the naked eye may look very similar, the processes by which they are and
were manufactured give them very different properties. As steel became better
understood and more efficiently produced in the early 1900s, wrought iron and
cast iron were mostly phased out as structural metals.
During the last half of the 19th Century, however, wrought iron was a
common structural material, and especially during the period of 1850-1880, many
bridges and buildings were constructed using wrought iron as a main structural
component. The invention of the grooved rolling process for wrought iron was the
critical development that allowed wrought iron to be produced in large quantities,
and used in the construction of large bridges and buildings. Before the invention
of grooved rolling mills, wrought iron and other metals, could only be rolled into
flat sheets, which limited the final shapes that could be produced.
The typical manufacture process of wrought iron in the 19th Century
consisted of, the melting of the pig iron in the hearth of a reverbatory furnace
which is lined with iron oxides, resulting in the elimination of most of the carbon,
5


silicon, manganese, phosphorus, and sulphur present in the charge by
oxidation (Mills & Hayward, 1922). The melted pig iron was then rolled in to bars
called muck bars which were subsequently bundled together and rolled again to
produce merchant bars. According to Adelbert Mills and Harrison Hayward in
their publication entitled Materials of Construction: Their Manufacture and
Properties,
This process results in the production of very pure iron mixed with from
1 % to 3% of slag, which the rolling process has caused to assume the
form of greatly elongated particles in the direction of rolling. This
circumstance accounts for the characteristic fibrous structure of wrought
iron (Mills & Hayward, 1922).
These merchant bars were also referred to as single-refined iron, describing the
single re-rolling of the original muck bars. In many instances, the merchant bars
would be piled together and rolled again, producing double-refined bars. This
repeated rolling of the iron produced, further elongation of the strands of slag in
the direction of rolling, thereby rendering the iron still more fibrous in its
structure (Mills & Hayward, 1922). As discussed in the Properties of Wrought
Iron section of this Thesis, the fibrous structure created by the iron silicate is a
unique and important feature of historic wrought iron.
Structural Applications of Wrought Iron
The continued improvement of the rolling process through the first half of
the 19th century enabled wrought iron to be used in increasingly wider
applications. Enhancements to the rolling process allowed wrought iron to be
6


rolled in various structural shapes. Additionally, during the middle of the 19th
Century, as the railroad industry was booming across the United States, many
bridges built of cast iron had collapsed, so builders began to turn to wrought iron
as a more ductile and reliable structural metal. During this period, while there
was a continual increase in the production and use of wrought iron, the first rolled
wrought iron structural shapes were produced.
While Ferdinand Zores of Paris is apparently the first person to have
successfully produced rolled wrought iron I-beams, Peter Cooper was the first to
roll structural shapes in the United States, at his Trenton Iron Works in New
Jersey. Although these processes enabled the rolling of wrought iron structural
shapes, the production of these shapes was still limited by the size of iron ingot
that could be processed in the furnaces of that time. Due to this limitation, large
wrought iron beams had to be created by the riveting together of smaller, more
manageable sections. Several buildings in the United States which used wrought
iron beams were constructed during the second half of the 19th Century. It was
not until the development of the Bessemer process for the refining of steel, as
well as the invention of elevators that buildings were able to be built taller than
six stories and be constructed of a steel skeleton, without the use of masonry,
allowing for more window space. These developments led to the beginning of the
skyscraper era in the first half of the 20th Century.
7


Properties of Wrought Iron
In the 19th Century, prior to the creation of governing bodies and
organizations with the authority to regulate the production of structural metals,
wrought iron, and other structural materials were commonly manufactured with
varying properties. In many cases the wrought iron produced by one refinery
would not have the same properties as that produced by another refinery.
Although lacking the strict standardization of modern structural metals, there are
characteristics that are common to historic wrought iron, regardless of where the
metal was produced.
The first, and perhaps most distinctive, characteristic of wrought iron is the
slag that is present in the metal. The definition of wrought iron according to the
American Society for Testing and Materials (ASTM) from 1930, states that
wrought iron is, a ferrous material, aggregated from a solidifying mass of pasty
particles of highly refined metallic iron, with which, without subsequent fusion, is
incorporated a minutely and uniformly distributed quantity of slag (Aston & Story,
1939). This quote is taken from a publication by James Aston and Edward B.
Story entitled Wrought Iron: Its Manufacture, Characteristics and Applications. In
this publication, Aston and Story go on to describe this slag that is unique to
historic wrought iron. They state that the, slag content in wrought iron may vary
from about 1% to 3% or more, by weight, and in well made wrought iron, there
may be 250,000 or more of these glass-like slag fibers to each cross-sectional
square inch (Aston & Story, 1939). Figure 2.1 shows a photo with 100X
magnification, clearly displaying the fibrous slag embedded in the base material.
8


As can be seen in Figure 2.2, even in an un-magnified photograph of a ruptured
specimen, the fibrous nature of the wrought iron is visible.
Figure 2.1: Wrought Iron Photomicrograph (Aston & Story, 1939)
Figure 2.2: Ruptured Wrought Iron Specimen
9


While steel has only a crystalline structure, wrought iron has, in addition to
its crystalline iron structure, a fibrous structure which is essential to the ductility of
the material, as well as its high resistance to fatigue and corrosion. However,
due to the single direction distribution of the slag fibers, caused by the rolling
process, wrought iron has a much higher tensile strength in the direction of rolling
than in the direction orthogonal to rolling. In other words, wrought iron has a
higher ultimate tensile strength (Fu) and yield strength (Fy) in the longitudinal
direction. Because the physical properties of wrought iron are largely those of
pure iron, the quantity and distribution of the fibrous slag has a great effect on
the physical properties of the wrought iron sample (Aston & Story, 1939). Figure
2.3 shows a compiled list of the strength of various wrought iron members. These
values were published in the catalogs of rolling mills during the last 30 years of
the 19th Century.
In addition to its high ductility compared to that of cast iron, wrought iron
was used extensively in the 19th Century, because it showed high resistance to
fatigue. According to Aston and Story,
Wrought iron is known to be relatively insensitive to notch effects and
unusually resistant to over stress. These desirable properties are
attributed primarily to high ductility, and of particular importance, to the
presence of the slag fibers which confer on the metal a tough, fibrous
structure somewhat analogous to that of a stranded wire cable (Aston &
Story, 1939).
They go on to state that, the slag fibers apparently serve to minimize stress
concentration and deflect the path of slip planes that develop in a metal under
the influence of conditions which ordinarily would result in fatigue failure (Aston
& Story, 1939). These desirable properties enabled wrought iron to be the
10


dominant structural metal for most of the 19th Century, until steel was able to be
produced with more reliable and desirable properties and in mass quantities. The
list shown in Figure 2.3 does not list any values for ductility or elongation percent
of the material. This is because through most of the 19th Century, engineers
were typically given only values of the yield and ultimate strength of a material. In
an article published in 2005 in the Journal of Materials in Civil Engineering
entitled Evaluation of Wrought Iron for Continued Service in Historic Bridges,
Robert Gordon and Robert Knopf state that, One of the earliest quantitative
ductility specifications was published in the 1873 account of the construction of
the International Bridge over the Niagara River at Buffalo, NY (Gordon & Knopf,
2005). Although historic wrought iron had a reputation for high ductility and
fatigue resistance, small variations in the chemical composition of wrought iron,
while not greatly affecting the ultimate strength, were known to have significant
effects on the metals ductility. Similar to the test results presented in this thesis
(see Chapters IV through VI) Gordon and Knopf discuss the wide range of
ductility found in tested wrought iron specimens from the 19th Century, As the
test data show, a wide range of ductilities can be attained in wrought iron, either
intentionally, or in the hands of less skilled makers, by accident (Gordon &
Knopf, 2005). The information presented by Gordon and Knopf describe methods
for estimating ductility of existing wrought iron structural members without
destructive testing.
11


Strength of Wrought Iron
for Use in Bridge Members
as Published in the Catalogs of Rolling Mills*
mr 1*73 Mfliimp wim Caraegic Kkxntn St Co ("Factor of Safer) J-)
1IT4 New Jersey Sled A Iron Co.
1*81- Carnegie Brother* St Co, Lid.
18*4
1U4 The Passaic Rolling Mill Co. foda(l-2dia.) 50,000-55,000
Urge ban (rtrolied) 46,000 47,000
plates uu) shape* ( im- 1*86 The Phoenix Iron Company 46,000 50,000
IMS- Pottxvillelron St Steel Co.
1*87.
1(88 rouods-(l*) 52,2tO
C2) sows
(4*1 4**220
1 1 JJ 51,000 49,500
(12*) 49,0*0
angles (3) 49,000
(6*) 49,160
beams flange* 51JM0
webs 50,130
1819 Camsgie Phipps St Co, Ltd.
1*92 Pencoyd Iron Works, A. P. Roberts & Co 50,000
189) Carnegie Steel Co.* Ltd. 48,000-50,000
plates (*"-24" wide) 44.000
angles and other shapes 48(000
pins 50.000
Hah Slrcssfpsil
14,(100
12,000
10,000
10.000
12,000
12.000
10,000
Figure 2.3: Wrought Iron Physical Properties (Aston & Story, 1939)
In addition to the physical properties of wrought iron, the chemical
properties of this unique structural metal must be understood in the context of the
inherent slag distribution. For chemical analysis of historic wrought iron, the two-
component nature of the metal must be taken into account. A thorough analysis
of wrought iron should provide the distribution of chemicals in each component,
both the slag and the iron. A typical wrought iron specimen will likely contain
small amounts of the following metalloids: carbon, phosphorus, manganese,
silicon and sulfur. Table 2.1 provides the chemical distribution of a typical wrought
iron sample. For comparison, Table 2.2, a chemical analysis from a typical early
12


20th Century mild steel, has been provided. As can be seen from these tables,
the carbon content of wrought iron (0.02% in sample shown here) is significantly
lower than that of steel, although in some cases, good wrought iron may have a
carbon content of 0.08% or 0.10% (Aston & Story, 1939). Higher amounts of
carbon in a wrought iron sample may create a higher strength wrought iron, but
will also significantly decrease the ductility of the material. A higher carbon
content than what is expected in wrought iron, may be an indication of imperfect
or incomplete refining, or may awaken suspicion that steel scrap has been used
in bushelling or piling [added during the forging of the iron] (Aston & Story,
1939). Another significant difference between historic wrought iron and early
steel is the manganese content. In the words of Aston and Story, The virtual
absence of manganese in wrought iron and its almost universal presence in steel
has resulted in the manganese determination being used as a means of
identification and differentiation (Aston & Story, 1939). The chemical phosphorus
has a significant impact on the ductility of the wrought iron; wrought iron with a
higher phosphorus content will display reduced ductility.
Seemingly slight variations in the chemical content of wrought iron,
incurred during the manufacturing stage, have the ability to produce metal with
greatly varied physical properties. For this reason, where possible, it is important
to perform a detailed chemical analysis of the historic wrought iron sample in
question. However, due to financial and scheduling limitations, a chemical
analysis was not performed on the material that has been tested for this study. A
simple spark test was performed to verify the wrought iron nature of the
13


material, as well as visual inspection. See Chapters IV VI for the configuration,
results and discussion of the spark test. Even without a chemical analysis,
conclusions can be drawn regarding the chemical content of the tested material
based on how the material performs under physical testing. See Chapters VI and
VII for this discussion and analysis.
Table 2.1: Typical Wrought Iron Chemical Analysis
(Adapted from Aston & Story, 1939)
Combined Analysis (%) Separate Analysis (%)
Base Metal Slag
Carbon 0.02 0.02 -
Manganese 0.03 0.01 0.02
Phosphorus 0.12 0.1 0.02
Sulphur 0.02 0.02 -
Silicon 0.15 0.01 0.14
Slag by Weight 3 - -
Table 2.2: Early 20th Century Mild Steel Chemical Analysis
Adapted from Aston & Story, 1939)
Metalloid Content (%)
Carbon 0.1
Manganese 0.5
Phosphorus 0.04
Sulphur 0.05
Silicon 0.1
14


Table 2.3: Results of Tensile Tests by Gordon and Knopf
(Adapted from Gordon & Knopf, 2005)
Yu= Upper Yield Stress, Yi= Lower Yield Stress, T = Ultimate Tensile Strength
r = Area Reduction, e = Elongation
Yu(ksi) Yi (ksi) T (ksi) r (%) e (%)
39.9 39.9 55.7 34 28
35.0 32.8 49.9 35 26
32.1 32.1 48.2 34 26
39.3 38.9 53.1 34 27
46.8 37.7 52.5 33 26
38.9 38.9 54.8 33 27
Average 38.7 36.7 52.4 33.8 26.7
Minnesota Bridge #5721
All the material used for the testing and analysis performed for this thesis
was donated by the Minnesota Department of Transportation (MnDOT), and was
taken from a bridge referred to as Minnesota Bridge #5721. This bridge, a 162
ft long Parker truss structure, was built in 1877 in the town of Sauk Centre, MN.
For 60 years it, enabled horses, wagons, buggies and pedestrians to cross the
Sauk River (Johnson, 2012). By the third decade of the 20th Century, the bridge
was no longer sufficient for the increased demand caused by the population
growth of Sauk Centre and rapidly growing automobile use. Because of this, in
1937, Minnesota Bridge #5721 was disassembled and in 1937 was re-erected in
a wilderness area in northern Minnesota, and for 70 years it carried logging
trucks and other vehicles at this scenic northern Minnesota location (Johnson,
2012).
15


After 70 years of service in its second location, Minnesota Bridge #5721
was once again deemed to be insufficient for its required use. Both the limited
size of the bridge, and its deteriorating structure made it clear that the bridge
could no longer be used for modern highway traffic. In 2009 the bridge was
removed from its second home and stored until 2011 in a MnDOT facility. After a
detailed structural study of the bridge, and a feasible rehabilitation plan was
completed, the bridge was installed in 2011 into its third and current location near
Stillwater, MN (see Figure 2.4). The renovations carried out on the bridge
included the replacement of the original deck with a lightweight concrete deck
and the replacement of deck stringers and floor end beams (originally wrought
iron), with new steel members. Additionally, eight of the 96 original wrought iron
eye-bar truss members were replaced with new steel members. The need to
replace eight of the original eye-bars was created during the disassembly of the
bridge in 2009 when, the disassembly contractors hardened tools produced
nicks and gouges in the heads of the eye-bars as the pinned connections were
taken apart (Johnson, 2012).
16


Figure 2.4: Minnesota Bridge 5721 Stillwater, MN
(Image used under Creative Commons Attribution-Share Alike 3.0 Unported
License)
Six of the eight eye-bars removed from the bridge were donated to the
University of Colorado Denver in 2012. The test specimens used for this study
were taken from two of the eye-bar sections that were removed during the latest
rehabilitation and relocation of Minnesota Bridge #5721.
Repair & Preservation of Structural Wrought Iron
As described in the previous section, since the dawn of the steel era,
modern manufacturing and modern construction methods, the need for new
structural wrought iron has mostly disappeared. However, because the wrought
iron manufactured in the late 19th Century was usually very durable, there are
still many structures in use today that rely on wrought iron as significant portions
of their structural systems. It is critical that engineers and contractors tasked with
the repair and rehabilitation of these structures be familiar with available and
17


proven repair methods. In this way, it can be ensured that these historic
structures last for many more years, not simply for their historic value but also for
their functional role in society. In recent years there has been some research
regarding the repair of historic wrought iron using modern construction methods.
However, even with the methods that have been studied, there is still a great
need for the further study of viable repair options, especially those utilizing
modern welding techniques. Chapters IV through VI of this thesis provide a
detailed analysis and testing results of the specific welded repair method studied
herein. The following sections in this chapter seek to summarize the prominent
historic wrought iron repair methods that have recently been studied by other
parties.
Testing of Historic Structural Wrought Iron
Throughout much of the 20th Century, many tests have been performed
on structural wrought iron which was manufactured during the 19th Century.
These tests include valuable information such as: detailed chemical analysis,
evaluation of the elastic modulus, determination of yield strength and ultimate
strength as well as measurements of percent elongation and toughness.
Examples of some of the available data are displayed in the section Properties
of Wrought Iron, of this thesis. The data shown there was all from material
produced in the late 19th and early 20th Centuries.
More recent test data can be found in the results published in the article
Evaluation of Wrought Iron for Continued Service in Historic Bridges. In this
18


study, Gordon and Knopf performed physical and chemical testing on wrought
iron samples from several historic bridge structures. Similar to the results
presented in Chapter V of this thesis, Gordon and Knopf report that,
Plastic deformation usually began with a sharp yield point followed by
discontinuous yielding that continued until steady work hardening took
over, leading to necking down and eventual ductile rupture of the
specimen (Gordon, 2005).
The results of their tension tests can be seen in Table 2.3. Gordon and Knopf
also performed significant chemical analysis on wrought iron samples.
Results from other tests performed on historic structural wrought iron can be
found in the following section of this thesis (Tables 2.4 and 2.5). Many other tests
have been performed over the last 100 years on historic structural wrought iron
from structures built in the 19th Century. Many of these tests show similar results
to the ones presented here and the data collected by this study. The discussion
of wrought iron testing by other parties in this paper has been limited to include a
small number of studies for the purpose of comparison and verifying the
legitimacy of the results obtained by this study.
Welding Repair of Structural Wrought Iron
To the knowledge of this author, only two previously published studies
present a thorough investigation of the ability of modern welding techniques to
repair historic structural wrought iron. The first is a study titled Arc Welding
Procedure for Repairing Wrought Iron in Historic Bridges, a Masters Thesis by
Preeda Chomsrimake at the University of Colorado Denver.
19


The study performed by Chomsrimake at the University of Colorado
Denver tested the ability of a single bevel groove weld to connect two separate
pieces of historic wrought iron. In Chomsrimakes study, 1/8 in. diameter, E7018
rods were used along with a groove made with a 45 degree included angle.
These welds were performed using a 1/4 in. by 1 in. steel backing bar, which was
removed prior to testing of the specimen (Chomsrimake, 2012). Figure 2.5 shows
a cross-section sketch of the weld joint used by Chomsrimake.
Figure 2.5: Single V-Groove Weld (Chomsrimake, 2012)
Chomsrimake performed tension tests and guided bend tests on the base metal
and the welded specimens and tabulated the results. The data from
Chomsrimakes tests showed the base metal specimens (no welds) to have an
average ultimate strength of 47.5 ksi and a average yield strength of 28.5 ksi.
The welded specimens that were subjected to the tension test demonstrated an
average ultimate strength of 46.6 ksi and an average yield strength of 28.0 ksi
(Chomsrimake, 2012). According to the author, this slight variance in values
20


between the the base metal specimens and the the welded specimens is not
significant and, Using SMAW to make V-groove welding in a butt joint of the
sample members still maintains the ultimate tension capacity of the historic
wrought iron members. (Chomsrimake, 2012). The average elongation reported
from these tension tests was 17.1% for the base metal specimens and 9.85% for
the welded specimens. These values show a significant decrease in the ductility
of the base metal specimen when compared to the welded specimen and,
according to the author, this is likely due to the higher yield strength and stiffness
of the weld metal compared to that of the wrought iron base metal.
(Chomsrimake, 2012). The summary of the tension test results performed by
Chomsrimake can be seen in Table 2.4.
Table 2.4: Results of Tensile Tests by Chomsrimake
(Adapted from Chomsrimake, 2012)
Fy (ksi) Fu (ksi) Elongation (%)
Base Metal Average 28.5 47.5 17.1
Base Metal St. Dev. 1 1.67 2.48
Welded Specimen Average 28 46.6 9.85
Welded Specimen St. Dev. 0.345 0.574 1.66
In addition to tension tests of the base metal and welded wrought iron
specimens, the 2012 study at the University of Colorado Denver also included
guided bend tests of both the root and the face of the V-groove weld. (Reference
Chapter IV for description of root and face guided bend tests.) According to
Chomsrimake, the root bend tests passed with only very minor cracks visible in
the welded zone. However, during the face bend test, the welded wrought iron
21


specimen was broken in the base metal near the welded joint (Chomsrimake,
2012).
The other existing study to investigate the ability of modern welding
techniques to repair historic structural wrought iron was performed at Purdue
University by Mark D. Bowman and Amy M. Piskorowski, entitled Evaluation and
Repair of Wrought Iron and Steel Structures in Indiana. In the Purdue study,
Bowman and Piskorowski performed tests using several wrought iron repair
methods, in an effort to develop a list of suggested repair techniques for joints
and members typical of wrought iron bridge constructions (Bowman, 2004).
Although they performed tests on many repair methods, only their tests on
welded wrought iron specimens will be discussed in this thesis. Bowman and
Piskorowski used a double bevel V-groove butt joint with an groove angle of 60
degrees and a root opening of 1/16 in. on a 1/2 in. thick wrought iron eye-bar
specimen. According to the authors, Weld passes were alternated on either side
[of the joint] to ensure that heat distortion in the piece was minimized (Bowman,
2004).
Regarding the results of the tension tests in general, the authors of the
Purdue study state that, All of the fractures of the historic wrought iron tensile
coupon testing were somewhat brittle in nature, and that, before any of the
specimens were about to fail during testing, there was no visible necking or any
typical pre-failure behavior that is typically found in structural steel (Bowman,
2004). Bowman and Piskorowski tested several tensile coupon specimens, five
of which were welded specimens, using the welding process described above.
22


The summary of their tensile test results is shown in Table 2.5 below. According
to the authors, none of the welded tensile tests failed in the region of the weld,
and there was little variation between the welded and non-welded specimens
when comparing both yield strength and ultimate tensile strength. From their
analysis and results, the authors of the Purdue study surmise that the welded
repair method they used can be considered a satisfactory weld detail (Bowman,
2004). Bowman and Piskorowski performed only tension tests on their welded
specimens, and as a result do not have any data on ability of the welded repair
method to resist bend tests. Nonetheless, they demonstrated that a specific
welding technique, using modern welding methods, can be used to sufficiently
repair a wrought iron member which is subject to tension only.
Table 2.5: Results of Tensile Tests by Bowman & Piskorowski
(Adapted from Bowman, 2004)
E (ksi) Fy (ksi) Fu (ksi) Elongation (%)
Average 27,700 31.6 47.2 11.7
St. Dev. 500 1.28 3.38 4.5
Although the tests and studies performed by other parties, and discussed
above, provide valuable information to engineers and those involved in the repair
of historic wrought iron structures, it is clear that there is ample need for more
studies and analysis, especially those involving the use of modern welding
techniques on historic structural wrought iron.
23


CHAPTER III
WELDING BACKGROUND
Welding Methods
Many modern welding methods have the potential to be useful for the
repair of historic structural wrought iron. Three of the most common modern
welding techniques will be discussed in the sections below, as they have the
highest potential to be used for the successful and economical repair of historic
wrought iron. Due to the limited scope of this study, this research investigated
only the ability of shielded metal arc welding (SMAW) for wrought iron repair. The
methods used, results and discussion of the SMAW process use for the repair of
historic structural wrought iron can be found in Chapters IV, V and VI.
Shielded Metal Arc Welding
The first, and perhaps most common modern welding technique,
especially for in-situ repairs, is the SMAW process. In the shielded metal arc
welding process, a flux-covered welding rod acts as the electrode and when
placed in proximity to the base metal, the rod forms an electric arc which creates
temperatures of 6500-7000 F. This heat melts both the base metal and the
welding rod, and as the rod is melted into the base metal the flux forms a slag
(not to be confused with the silica slag present in historic wrought iron) which
floats to the top of the weld and hardens as the metal cools. This slag is removed
24


after each weld pass. With the welding rod acting as both the filler metal and the
electrode, SMAW is a simple and easy welding process.
The SMAW process can be performed with alternating current (AC) or
direct current (DC) power. However, for the research discussed here only DC
power was used as DC power is more commonly used in construction
applications. Additionally, when DC power is used, with the SMAW process, a
welder has the option of using direct current electrode positive (DCEP) or direct
current electrode negative (DCEN). These two types of DC welding simply refer
to the end of the circuit on which the electrode is placed. For DCEP, the electrons
in the circuit flow from the positive pole of the welding machine to the electrode,
and the negative pole of the machine is attached to the base metal to complete
the circuit. For a DCEN setup, the electrons in the circuit flow from the negative
pole of the welding machine to the electrode, and the positive pole of the
machine is attached to the base metal to complete the circuit. Typically the
decision to use DCEP or DCEN is driven by the welding rod selected for the job,
as each rod has certain limitations that govern the welding circuit type on which it
can be used.
The selection of the type of welding rod is crucial for a successful weld.
The choice of welding rod can be based on many factors, including base metal
composition, joint type, base metal thickness, and welding environment. For this
study, the common and readily available E7018 rod was used. The nomenclature
used to designate welding rods for the SMAW process is as follows: the first two
numbers, in this case 70, refer to the minimum required tensile strength of the
25


filler metal in k/in2 (ksi). The third number in the rod designation specifies the
recommended welding position for that rod type (flat, horizontal, vertical, etc.).
The fourth number in a typical rod designation specifies the composition of the
welding rod covering, or flux, and the suggested power supply for that rod
(DCEP, DCEN or AC). These final two numbers should be read together to inform
the welder if the chosen rod is the best fit for the application. In the case of an
E7018 rod, the 18 indicates a covering composition containing iron powder, low-
hydrogen and potassium, as well as indicating that the rod can be used in all
positions, and with DCEP or AC power supply. In addition to the 4 standard
numbers in a welding rod designation, a suffix can be added to describe a low-
alloy steel electrode. In this case, the additional suffix provides information
pertaining to the composition of the filler metal. For example, the suffix -A1
indicates the steel rod is a carbon-molybdenum electrode containing 0.5%
molybdenum. Table 3.1 lists the available carbon steel covered electrodes
defined by the American Welding Society (AWS) with the tensile strengths of 60
and 70 ksi (E60XX and E70XX). This information has been adapted from AWS
A5.1-91 (AWS, 1991). The information presented in Table 3.1 includes only those
electrodes which have a consumable metal tensile strength of 60 and 70 ksi, as
these are the most commonly used electrodes for structural applications. For the
repair of historic wrought iron, the use of higher strength electrodes is not
necessary, as the strength of the electrode will always exceed the strength of the
wrought iron base metal which typically has a tensile strength in the range of
26


45-55 ksi (Aston & Story, 1939), and is consistent with the testing performed as
part of this thesis.
Table 3.1: Carbon Steel Covered Arc Welding Electrodes
Adapted from AWS A5.1-91 (F: Flat, V: Vertical, OH: Over-head, H: Horizontal)
AWS Classification Flux Type Recommended Position Electric Current Type
E6010 High cellulose sodium F, V, OH, H DCEP
E6011 High cellulose potassium F, V, OH, H AC or DCEP
E6012 High titania sodium F, V, OH, H AC or DCEN
E6013 High titania potassium F, V, OH, H AC, DCEP, DCEN
E6020 High iron oxide H-fillets AC or DCEN
E6022 High iron oxide F AC, DCEP, DCEN
E6027 High iron oxide, iron powder H-fillets, F AC or DCEN
E7014 Iron powder, titania F, V, OH, H AC, DCEP, DCEN
E7015 Low hydrogen sodium F, V, OH, H DCEP
E7016 Low hydrogen potassium F, V, OH, H AC or DCEP
E7018 Low hydrogen potassium, iron powder F, V, OH, H AC or DCEP
E7024 Iron powder, titania H-fillets, F AC, DCEP, DCEN
E7027 High iron oxide, iron powder H-fillets, F AC or DCEN
E7028 Low hydrogen potassium, iron powder H-fillets, F AC or DCEP
E7048 Low hydrogen potassium, iron powder F, OH, H, V- down AC or DCEP
27


Gas Metal Arc Welding
Another common and fairly simple welding method is the Gas Metal Arc
Welding (GMAW) process. In the GMAW process, a continuously fed wire
provides the consumable electrode that creates the arc and is melted into the
weld. To shield the weld and protect it from impurities, the welding arc is shielded
by an inert gas. This process is also commonly known as metal inert gas (MIG)
welding, and is used with a constant voltage DC welding machine. Commonly,
gases such as argon, carbon dioxide or helium are used as the shielding gas,
and are selected for use based on the type of base metal being welded. For the
GMAW process, a single insulated cable caries the wire feed as well as the feed
for the shielding gas.
The GMAW process has potential to be used for wrought iron repair,
because for a typical groove weld, a GMAW weld requires a much smaller
groove angle for a quality penetration weld. Because of this, less filler metal is
required, and less heat is transferred to the base metal. Due to its simplicity,
GMAW is one of the most efficient welding processes, and with multiple wire
types and shielding gas types available for use, it provides great potential for the
successful repair of historic wrought iron structures.
28


Gas Tungsten Arc Welding
Similar to the GMAW process discussed above, the gas tungsten arc
welding (GTAW) process uses a shielding gas to prevent impurities in the weld.
However, in the GTAW process a non-consumable tungsten electrode is used, as
opposed to the consumable electrodes of the SMAW and GMAW processes. In
this process a filler metal may or may not be used, and an AC or DC power
source may be used. The GTAW process is commonly referred to as tungsten
inert gas (TIG) welding. Inert gases that are commonly used with the GTAW
process include, but are not limited to, Argon, Helium, and Argon-Hydrogen
mixtures. Compared to the SMAW and GMAW processes, the GTAW process is
more difficult to perform correctly, but when properly executed produces
extremely clean, high quality welds. For this reason, the GTAW process is used
frequently in the welding of pipe applications such as petroleum and chemical
pipe lines. Although technically more difficult, the GTAW process has potential to
be used to perform high quality repair of historic structural wrought iron.
Standard Joint Types
In modern welding practice there are five standard joint types that are
commonly used. Although very limited research has been performed using
modern welding techniques on historic wrought iron, all five of the standard joint
types have the potential to be used for the repair of historic wrought iron
structures, and will be presented below, in hopes that future research will
performed using all these joint types with wrought iron. The five standard joint
29


types are as follows: butt joint, corner joint, t-joint, lap joint and edge joint. Figure
3.1 is a graphical representation of the five standard modern welding joint types.
Based on the application of each of these joints, and the required strength of the
welded joint, several types of welds can be applied to each of these joint types.
For some weld types, such as the double-bevel groove weld performed on a butt
joint, special preparation of the joint is necessary. For other weld types, such as a
fillet weld performed on a t-joint, no unique joint preparation is required. Table 3.2
shows an abridged list of various welds that can be performed on each joint type.
For this study, the research focused on developing an acceptable welding
procedure using a butt joint and a double bevel groove weld (see Figure 4.8).
This was the only joint utilized for this research, due to the necessity of changing
only one variable at a time, and the iterative nature of this process. See Chapter
IV for a description of the welding processes performed in this study.
BUTT JOINT T-JOINT |__|_
EDGE JOINT
-* LAP JOINT
CORNER JOINT
Figure 3.1: Standard Joint Types
30


Table 3.2: Applicable Welds for Basic Joint Types
Adapted from AWS A3.0: 2001
Joint Type Applicable Welds
Butt Joint Bevel-groove, Flare-bevel-groove, Flare-V-groove, J-groove, Square-groove, U-groove, V-groove
T-Joint Fillet, Bevel-groove, Flare-bevel-groove, J-groove, Square- groove
Edge Joint Bevel-groove, Flare-bevel-groove, Flare-V-groove, J-groove, Square-groove, U-groove, V-groove, Edge
Corner Joint Fillet, Bevel-groove, Flare-bevel-groove, Flare-V-groove, J- groove, Square-groove, U-groove, V-groove
La Joint Bevel"9roove> Flare-bevel-groove, J-groove, Square- p groove
Weld Quality, Flaws & Testing
Repairing historic structural wrought iron with modern welding methods
has great potential, but while researching potential welded repair methods, it is
imperative that engineers understand the effect that even very small defects can
have on welds. The following discussion will be limited to the common defects
and problems that are common to welds performed using the SMAW process, as
that was the only process used for the research described by this thesis.
As described above, electrodes used for shielded metal arc welding are
coated with a flux that, when melted during the welding process, floats to the top
of the molten weld puddle and hardens as the metal cools. This coating that ends
up on top of the welded metal is called slag. For all multi-pass welds, such as
those used for the double bevel welds described in Chapter IV of this thesis, the
slag must be completely removed with a wire brush and/or chipping hammer
31


before the next weld is performed. Any remaining slag that is left can create a
slag inclusion, which will leave pockets in the weld, and can significantly
decrease the strength of the welded joint. Slag inclusions are likely to leave the
joint itself weaker than the base metal, which in structural practice will create a
potential stress concentration that may lead to a sudden failure of the joint.
Beyond the easily avoided problem of slag inclusion, there are several
factors that will influence the quality of a weld using the SMAW process. Four
significant factors that are controlled by the welder are, arc length, welding travel
speed, width of weld bead, angle and position of electrode. Arc length refers to
the distance between the electrode and the base metal, and affects the heat of
the weld, the quality of penetration and the purity of the weld. Welding travel
speed is the pace at which the welder moves the electrode from one end of the
weld to the other. Welding speed that is too fast or slow can significantly
decrease the quality of the weld by causing various defects. The width of the
weld bead is varied by the motion of the electrode and the speed of travel. The
correct width of the weld must be applied by the welder to ensure that sufficient
penetration is provided, but also that excessive heat is not applied to the base
metal. The angle of the electrode refers to how the electrode is positioned with
respect to the face of the base metal, and can affect the penetration of the weld,
and the formation of a proper weld bead.
All of these factors must be balanced and adjusted appropriately by the
welder based on the joint type selected, the base metal, the electrode being used
and the type of weld required. If the correct adjustment is not made, several weld
32


defects, or flaws can occur. Some of these weld defects are easily visible to the
naked eye. However, some defects are only discovered through the use of more
sophisticated testing methods.
Four common defects that can be identified by simple visual inspection
are poor welding proportions, undercutting, poor penetration, and surface flaws
and defects. An experienced welder can not only recognize all of these defects
but also recommend appropriate adjustments to the welding process to fix them.
Flaws and defects that cannot be detected by visual inspection will only be found
by destructive and non-destructive testing. Two non-destructive testing methods
that are common in the construction industry are ultrasonic and x-ray testing.
Both of these methods are used to the verify the quality of welds made in the
construction industry and have the potential to be used to investigate welds
performed on historic wrought iron. Additionally there are several destructive
testing methods used to test the quality of a weld, and that should be used while
performing research on welding methods for historic wrought iron. The two most
common, and the two used for the research presented by this thesis, are the
tensile test and the guided bend test. For a detailed description of these tests see
Chapters IV-VI.
Welding Procedure Qualification
Modern welding procedures used in the structural engineering and
construction industry today are governed by the American Welding Society
(AWS) Structural Welding Code, AWS D1.1. This standard includes many pre-
33


qualified welding procedures intended for use in structural steel construction.
Although many of these pre-qualified welding procedures are available for the
welding of structural steel, in researching the available literature, the author has
found no published, pre-qualified welding procedure for the welding of historic
structural wrought iron. The purpose of the research discussed in this thesis is to
provide a tested welding procedure to be used on historic wrought iron. My hope
is that the welding procedures discussed in this thesis, or similar procedures will
eventually be put through an AWS approved welding procedure qualification
testing program.
The essential variables that must be defined in a pre-qualified welding
procedure, and that are applicable to the SMAW process, are the following: base
metal thickness, base metal properties, filler metal properties, pre-heat
temperature, post-heat temperature, joint geometry, and welding position. In
many structural applications, if a type of weld is required that does not fit within a
pre-qualified welding procedure, a welding procedure qualification record
(WPQR) must be submitted and approved prior to the weld being used in a
fabrication shop or in the construction field. For the qualification of a groove weld,
specimens must be cut from the welded material and used for tension tests and
guided face and root bend tests to verify the strength, and ductility of the final
welded specimen. These tests were performed on the welded wrought iron
studied for this thesis, and the testing methods and results can be found in
Chapters IV-VI.
34


CHAPTER IV
WROUGHT IRON TESTING TEST DESIGN & CONFIGURATION
Base Metal Testing Design & Configuration
Test with 220k MTS
Before any tests on welded wrought iron were performed, and before any
welds were completed, samples of the non-welded wrought iron obtained from
Minnesota Bridge #5721 were tested in order to determine basic material
physical properties. For this reason, tensile coupon tests were performed on five
base metal specimens. Three of these tests were performed on a 220k MTS
testing frame, shown in Figure 4.1. The other two utilized a 20k MTS testing
frame, which can be seen in Figure 4.2. Both testing machines are located in the
Civil Engineering Laboratory at the University of Colorado Denver. Initially, only
the larger of the two testing frames was used, in order to reduce the amount of
machining required of each test specimen. The 20k frame was used to verify the
results of the larger testing frame and to confirm the accuracy of all obtained
results.
35


Figure 4.1: 220k MTS Testing Frame
Figure 4.2: 20k MTS Testing Frame
36


For wrought iron samples to be tested using the 220k MTS, a system had
to be devised to mount the wrought iron specimen into the testing machine.
Unlike the 20k MTS which uses simple clamps to grip the specimen, the 220k
MTS utilizes a two inch diameter pin at each head to secure the specimen being
tested. The wrought iron material being tested was taken from two different eye-
bar sections of Minnesota Bridge #5721. The portions of these eye-bars that
were cut out and tested were rectangular in cross section, with approximate
dimensions of 7/8 in. thick and 2 in. wide. Because these specimens were limited
to a width of 2 inches, the 2 inch diameter pin of the 220k MTS could not be
directly connected to the wrought iron specimen. To solve this problem, two
clamp plates for each head of the MTS were machined from mild steel. Each
clamp plate contained a hole cut to fit the 2 in. pin, a 5/16 in. deep by 2 in. wide
slot to accept the wrought iron flat bar specimen, and two 13/16 in. diameter
holes to accept the 3/4 in. bolts which were used to clamp the two plates onto the
specimen, thus securing the wrought iron specimen through a clamping, friction
force. Figure 4.3 below shows the design of the mild steel clamp bars that were
fabricated for the purpose of testing the wrought iron samples with the 220k MTS
testing frame. To ensure sufficient friction force, and avoid transferring the load of
the test into the holes in the wrought iron specimen through bearing, 3/4 in.
diameter bolts were used, and a torque of 425 ft-lb was applied to each bolt.
37


Figure 4.3: Design of Steel Clamp Bar for Tension Test
The decision to rely on the friction force of the clamp plates was driven by
the inability of holes drilled in the wrought iron specimen to withstand the
required force through direct bearing action. Limited by the thickness and width
of the provided wrought iron samples, had the researchers relied solely on the
bearing strength of holes in the wrought iron, the allowable tension force in the
specimen would have been significantly decreased. Using the clamp plate
method and relying on the friction force to resist the tension loading from the
MTS machine allowed a higher force to be applied and greatly decreased the
38


required machining of the wrought iron specimen, thus expediting the research
process.
The wrought iron specimens to be tested were machined to have a
smaller cross-sectional area through the middle of the specimen, so the location
of failure in the specimen could be limited to a relatively small zone. A cross-
sectional area of approximately 0.75 in2 was obtained in the failure zone of the
specimen. The cross-sectional area in the failure zone was specified based on
the expected ultimate tensile strength of historic structural wrought iron (see
Chapter II of this thesis), and a preferred specimen failure in the range of 38,000
to 40,000 pounds of tensile loading. Each wrought iron specimen was cut to be
18 in. long, with a pair of holes drilled at each end to match the holes in the
clamp plate assembly. Figure 4.4 shows the design of the wrought iron base
metal specimens used in the 220k MTS test frame.
For the base metal tension tests, strain gauges were placed in the
expected failure zone. These gauges were used for the sole purpose of obtaining
an accurate stress-strain diagram, and thereby determining the elastic modulus
of the wrought iron material in question. The strain gauge data was collected in
tandem with data from the MTS machine until the yielding of the wrought iron
specimen occurred, at which point the strain gauges failed and stopped
producing usable data.
39


Figure 4.4: Design of Wrought Iron Tension Test Specimen
In addition to the strain gauges, gauge marks were placed on the wrought
iron specimen (see Figure 4.4) and the distance between gauge marks was
verified with a caliper to the nearest thousandth of an inch. After each specimen
was tested to failure, these gauges marks were measured a second time. The
difference between the first and second measurement of the gauge marks
provided an accurate measurement of the elongation of the specimen, which
40


allowed the researchers to calculate the ductility of each tested specimen. Figure
4.5 shows a picture of the complete base metal tension test, with the 220k MTS,
steel clamp plates, wrought iron specimen and strain gauges.
Figure 4.5: Base Metal Tension Test Setup
The tension tests of the base metal specimens were performed using a
displacement controlled testing function. The initial tests were performed at a rate
of 1/16 in. per minute, but after review of the initial data, it was determined that
there would be no negative affects if the rate was increased, and it would greatly
expedite the testing process. After the initial test, each tension test was
41


performed at a rate of 1/8 in. per minute. The computer console used for the
control of the 220k MTS was set to provide output data sets of displacement vs.
time and force vs. time.
Test with 20k MTS
The setup for the tests performed on the 20k MTS test frame was very
simple because of the configuration of the testing apparatus. As can be seen in
Figure 4.2, the 20k MTS has built-in clamps at each head. These clamps are
designed with a system of springs which increases the clamping force as the
heads move apart. Therefore, as the tension load on the specimen increases, the
friction force holding the specimen also increases.
The two specimens created for use with the 20k MTS were designed to
fail at approximately 15 k, allowing sufficient tolerance to ensure the specimen
failed within the range of the testing machine. To ensure failure at this lower
tensile load, the specimens were machined to have a rectangular cross-section
3/8 in. thick and 3/4 in. wide. This provided a cross-sectional area of 0.281 in2
throughout the failure zone. In addition to the displacement and force data
recorded from the heads of the testing machine, an extensometer (see Figure
4.6) was placed in the middle of the expected failure zone and was connected to
the MTS control computer in order to record strain during the elastic phase of
each test. The extensometer was removed prior to yielding of the specimen, in
order to prevent damage to the sensitive electronic instruments within the device.
42


The tests performed on the 20k MTS used the same set of parameters
which were used for the tests performed on the 220k MTS. Each of the two tests
was performed at a rate of 1/8 in. per minute. The computer console used for the
control of the 20k MTS was set to provide output data sets of time vs.
displacement and time vs. force, in addition to the output of strain values during
the elastic loading phase.
Figure 4.6: Wrought Iron Specimen for 20k MTS with Extensometer
The base metal tension tests demonstrated the physical properties of the
wrought iron material being studied and were essential to the later testing of
welded wrought iron specimens. From the base metal tension tests, the
researchers were able to determine the average elastic modulus, yield strength,
ultimate strength and ductility of the wrought iron material that is discussed in this
thesis. For the results of the base metal tension test, see Chapter V, and for a
discussion and analysis, see Chapter VI.
43


Base Metal Spark Test Design & Configuration
In addition to the tension tests described above, the wrought iron
specimens taken from eye-bars of Minnesota Bridge #5721 were subjected a
spark test. Simple in nature, the spark test is a quick, in-expensive way of
identifying the material at hand. Besides the characteristics demonstrated during
loading and failure, and visual examination of the failed specimens, the spark test
provides a third way to verify that the material being tested is wrought iron and
not another metal that has been mis-labeled.
The spark test was performed simply by taking a piece of wrought iron that
had been cut from one of the Minnesota eye-bars and grinding it on a standard
machine shop grinding wheel. In order to document the results of the spark test,
the lights were turned off, and photographs were taken while the wrought iron
sample was being pressed to the grinder and sparks were being generated. For
the results and discussion of the findings of the spark test see Chapters V and VI
of this thesis.
Welded Specimen Tension Tests Design & Configuration
Prior to the beginning of the welding process, a detailed welding plan had
to be developed. The goal of the welding plan was to control as many variables
as possible in the welding process, and change only one variable at a time,
thereby making it easier after testing to draw conclusions regarding which
welding procedures showed better performance. All welds were performed at
Western Colorado Community College, Grand Junction, CO, by an advanced
44


welding student. An eye-bar labeled M1 was used for all the welded tension
tests. This eye-bar was originally referred to as diagonal member U2W-L3W
while it was part of Minnesota Bridge #5721. Additionally, three specimens were
cut from eye-bar M5 or diagonal member U2E-L3E.
The welded specimens for the tensile tests were prepared by cutting 18.8
in. and 20 in. long sections out of the M1 and M5 eye-bars. Figure 4.7 shows the
diagram that was used to direct the cutting of eye-bar M1 in order to maximize
the number of usable specimens that could be obtained from the eye-bar. See
Figure 4.8 for a diagram of the segmentation of eye-bar M5.
Figure 4.7: Eye-Bar M1 Segmentation Diagram
Each of these sections (M1-1 through M1-9, and M5-10 through M5-12)
were cut in half and each half was prepared with a double bevel on one end. The
diagram in Figure 4.8 shows the type of weld that was used to weld together the
two halves of each specimen. For all specimens, the dimensions of the landing,
45


f and the root opening, FT were held constant at 1/16 in. The groove angles, a
and 3, were varied equally so that the top groove angle was always equal to the
bottom groove angle. In addition to the groove angle, the pre-heat temperature
was adjusted for each weld. Both the groove angle and the pre-heat temperature
were adjusted based on the welding parameters shown in Table 4.1. For each of
these welds, a Lincoln Electric Precision TIG 185 machine (see Figure 4.9) set
up for the SMAW, DCEP process was used along with 1/8 in. diameter, E7018
welding rods. The amperage was kept within a range of 115-130 amps, but within
that range, it was adjusted at the discretion of the welder (see Table 4.1).
Table 4.1: Weld Parameters for Welded Tension Tests
* Value not provided
Bar Number Angle (degrees, a = 3) Preheat Temp. (F) Amperage (amps)
M1-1 30 No Preheat 115
M1-2 30 300 120
M1-3 30 600 120
M1-4 45 No Preheat 130
M1-5 45 300 130
M1-6 45 600 125
M1-7 60 No Preheat 125
M1-8 60 300 125
M1-9 60 600 125
M5-10 75 No Preheat *
M5-11 75 300 *
M5-12 75 600 *
46


Figure 4.9: Lincoln Electric Precision TIG 185
After the completion of the welds on specimens M1-1 through M1-9, each
specimen was machined per the sketch shown in Figure 4.4, in order to be used
47


with the clamp plates in the 220k MTS testing frame. Each specimen was tested
using the same parameters used for the base metal tension tests, described at
the beginning of Chapter IV. Chapters V and VI contain the results, analysis and
discussion of the tension tests performed on the welded specimens.
Welded Specimen Bend Tests Design & Configuration
The welded specimens used for bend tests were created with a very
similar process to those used for the tension tests. For the bend tests specimens,
diagonal member U2E-L3E from Minnesota Bridge #5721, re-labeled as M5, was
used. This member was cut into several 8 in. long pieces (see Figure 4.10). Each
piece, labeled M5-1 through M5-9, was then cut in half and one end of each half
was prepared for the double bevel weld shown in Figure 4.8.
94" ()
M5-10T M5-11T M5-12T SPARE ^ P
1 1 1 1 1 1 1
18.8" 18.8" 18.8 18.8" 1 8 1
18.8
94" () j t-
80"
O .II I
' £>' '
I I I
\v Jo' Jo' Jp' Jp' Jo' Jo' Jp' Jo' Jo' Jp' Jo Jo
^ ^ ^ y ^ y y y y y
NOT USED
4" 8"
20
8 4" [ 8" 8" ,4"| 8" I 8 4 8 I 8"
20 T T 20" r T J 20"
Figure 4.10: Eye-Bar M5 Segmentation Diagram
48


After the completion of the weld, the 8 in. long welded specimen was cut
in half longitudinally, and machined to the dimensions required for the bend test
apparatus. The required preparation for the final specimen to be subjected to the
bend test is shown in Figure 4.11. This final cut and machining provided the
researches with two specimens (each with a thickness of approximately 3/8 in)
for each weld parameter listed in Table 4.2. This allowed a root and a face bend
test to be performed for each set of weld parameters.
Table 4.2: Weld Parameters for Welded Bend Tests
Bar Number Angle (degrees, a = 3) Preheat Temp. (F)
M5-1 30 No Preheat
M5-2 30 300
M5-3 30 600
M5-4 45 No Preheat
M5-5 45 300
M5-6 45 600
M5-7 60 No Preheat
M5-8 60 300
M5-9 60 600
49


For the root bend test, the the testing plunger is placed against the face of
the weld, so the side of the specimen with the root of the weld is subjected to
tensile stress. Similarly, for the face bend test, the specimen is oriented so the
face of the weld is subjected to tensile stress. A photograph of a final prepared
specimen, prior to bending can be seen in Figure 4.12. Provisions were made to
test the welded specimens using a root bend test and a face bend test, because
both tests are required to verify the quality of the weld and for welder
qualification.
50


After all the specimens were prepared and ready for testing, they were
transported to the welding shop and testing facilities of Emily Griffith Opportunity
School in downtown Denver. The bend tests were performed per testing standard
ASTM E290, Standard Test Methods for Bend Testing of Material for Ductility.
The guided-bend test, per ASTM E290, section 3.6 was used (ASTM E290-97a),
and the testing machine was set up to accommodate the 3/8 in. thick specimens.
The goal was for each specimen to be bent 180 degrees, or a Type 1 bend as
designated by ASTM E290. Figure 4.13 shows the schematic of the guided-bend
test that was used for this study. For the bend tests recorded in this thesis, a
testing machine equipped with a hydraulic plunger was used. The testing
machine used was the Watts W-50 Tensile & Bend Weld Tester, manufactured by
51


Watts Specialties of Milton, WA. See Figure 4.14 for a photograph of the bend
test machine which was used for this study.
The results, analysis and discussion of these bend tests can be found in
Chapters V and VI of this thesis.
C = D + 3T
D = PLUNGER DIAMETER
T = SPECIMEN THICKNESS
W = SPECIMEN WIDTH
Figure 4.13: Guided Bend Test Schematic
(Adapted from ASTM E290-97a)
52


Figure 4.14: Watts W-50 Tensile & Bend Weld Tester
53


CHAPTER V
WROUGHT IRON TESTING RESULTS
Base Metal Testing Results
As stated in the previous chapter, five base metal samples taken from
Minnesota Bridge 5721 were tested. Three base metal samples were tested on
the 220k MTS testing frame and two base metal samples were tested using the
20k MTS testing frame. These tests were performed in order to determine the
physical properties of the wrought iron material, as well as to provide a basis for
comparison with the welded wrought iron specimens. These specimens were
taken from eye-bar sections labeled M1 and M3 (see Chapter IV), and were
given the the label B-MX-X, designating them as base metal, not welded,
specimens. The following properties of the base metal were determined from
each test: yield strength, ultimate tensile strength, ductility (measured by percent
elongation) and percent area reduction (measured at location of specimen
failure). Additionally from one of the base metal specimen tests (B-M1-2) the
elastic modulus was determined from the stress-strain diagram. The elastic
modulus was not calculated for every specimen, due to unsuccessful
coordination of the strain gauge data with the MTS output data during some of
the tests. However, the small variance of the ultimate strength and yield strength
between the specimens suggests a consistent elastic modulus for all specimens.
Table 5.1 displays the results of the base metal tension tests. Graphical results of
test specimens B-M1-1 and B-M1-2 can be seen in Figures 5.1 and 5.2,
54


respectively. The remaining specimen graphical results can be seen in Appendix
A, and relevant photographs can be found in Appendix B. See Chapter VI for a
discussion of these test results.
Table 5.1 Base Metal Tension Test Results
Indicates value not calculated, or determined to be inaccurate
Specimen # Fy (ksi) Fu (ksi) % Elongation % Area Reduction E (ksi)
B-M1-1 34.3 49.6 N/A* 35.8 N/A*
B-M1-2 37.3 48.7 33 40.6 28,100
B-M3-2 37.3 50.7 32 N/A* N/A*
B-M3-3 39.1 56.4 22.5 N/A* N/A*
B-M3-4 33.6 47.4 28.1 42.4 N/A*
Mean 36.3 50.6 28.9 39.6 -
Std. Dev. 2.30 3.48 4.76 2.79 -
CV 0.06 0.07 0.16 0.07 -
55


FORCE (LB)
40000
36000
32000
28000
24000
20000
16000
12000
8000
4000
0


/
r /





i
0 0 1 0 2 0. 3 0 4 0 5 0 6 0 7 0 8 0 9 1.
DISPLACEMENT (IN)
Figure 5.1: B-M1-1 Force vs Displacement Plot
56


Stress (psi)
40000.0
30000.0
20000.0
10000.0
0.0023
0.0045
Strain (in/in)
0.0068
0.0090
Figure 5.2: B-M1-2 Stress vs Strain Plot
(Data collection stopped after yielding of specimen)
57


Spark Test Results
The spark test, performed to verify the wrought iron nature of the tested
material was performed using the shop grinding wheel at the University of
Colorado Denver. Because this is a visual test, the results are presented here in
photograph form in Figure 5.3. For comparison, a photograph of a spark test
performed on mild steel has been included (see Figure 5.4). Reference Chapter
VI for a discussion of these results.
Figure 5.3: Spark Test Photograph Wrought Iron from MN Bridge 5721
Note: Sparks appear as spurts, without forks or fingers, indicating very low
carbon content of metal
58


Figure 5.4: Spark Test Photograph Mild Steel
(Photo by Chomsrimake. Used with permission)
Note: Sparks fork at ends, indicating higher carbon content
Welded Specimen Tension Tests Results
The welded wrought iron specimens were tested using only the 220k MTS
testing frame with the process described in Chapter IV. The goals for the tension
tests on the welded specimens were two-fold; first to verify that the specimen
would fail well away from the heat affected zone (HAZ), and secondly to see if
the welding of the wrought iron would significantly affect the physical properties
of the metal. To this end, the same values shown for the base metal specimens
were determined for the welded specimens. Additionally, photographs are
included (Figures 5.5 and 5.6) to show the location of failure relative to the
location of the weld. In both photographs shown below it is clear that the
59


specimen failed away from the heat affected zone, and as shown in Table 5.2,
none of the nine welded specimens failed in the HAZ. The area reduction and
elongation, or necking, typical to all the tested specimens studied in this thesis,
can be seen by the photographs in Figures 5.5 and 5.6. This necking occurred in
the final minutes of loading prior to failure, after the yielding of the specimen. It is
of interest to note that many of the welded specimens experienced this necking
phenomenon on either side of the weld, and not just in a single location (see
Figure 5.6). See Appendix A for displacement vs force of all the welded
specimen tension tests. Additional photographs can be found in Appendix B.
Tables 5.3 and 5.4 show the average yield strength, ultimate strength and
percent elongation organized based on the parameters of the weld that were
adjusted during the research process: groove angle and preheat temperature.
The discussion pertaining to the results of the welded specimen tension tests can
be found in Chapter VI. Reference Table 4.1 for associated groove angle and
preheat.
60


Table 5.2 Welded Specimen Tension Test Results
Specimen # Fy (ksi) Fu (ksi) % Elongation % Area Reduction Failure in HAZ
M1-1 32.8 48.5 26 38.5 No
M1-2 32.7 50 23.8 25.8 No
M1-3 28.4 45.1 26.5 24 No
M1-4 31.1 50.1 24.5 24.9 No
M1-5 33.1 51.5 25.7 26.5 No
M1-6 33.0 50.7 25.6 26.7 No
M1-7 34.7 52.7 25.7 27.2 No
M1-8 31.8 48.6 24.9 24.5 No
M1-9 32.6 50.7 26.3 25.8 No
M5-10 30.8 48.1 24.0 36.8 No
M5-11 31.4 49.4 23.9 34.7 No
M5-12 32.6 48.2 21.1 37.4 No
Mean 32.1 49.5 24.8 29.4 -
Std. Dev. 1.56 1.97 1.51 5.64 -
CV 0.05 0.04 0.06 0.19 -
Table 5.3 Welded Specimen Tension Test Results By Groove Angle
(Values are averages of all welds performed with noted groove angle)
Groove Angle Fy (ksi) Fu(ksi) % Elongation
30 31.3 47.9 25.4
45 32.4 50.8 25.3
60 33.0 50.7 25.6
Std. Dev. 0.877 1.646 0.184
CV 0.027 0.033 0.007
61


Table 5.4 Welded Specimen Tension Test Results By Preheat Temperature
(Values are averages of all welds performed with noted preheat temperature)
Preheat Temp. (F) Fy(ksi) Fu (ksi) % Elongation
None Room Temp. 32.9 50.4 25.4
300 32.5 50.0 24.8
600 31.3 48.8 26.1
Std. Dev. 0.806 0.833 0.668
CV 0.025 0.017 0.026
Figure 5.5: Welded Specimen Tension Test Photograph 1
62


Figure 5.6: Welded Specimen Tension Test Photograph 2
Welded Specimen Bend Tests Results
In addition to the tension tests, guided bend tests were performed on nine
welded specimens. Each weld type was tested by a face bend and a root bend
test (see Chapter IV). The goal was for each specimen to be bent 180 degrees
without significant visual faults such as cracking or splitting. Of the nine root bend
tests that were performed, eight were successful, while one specimen broke well
before reaching the desired 180 degree bend. The face bend tests were not as
successful: all nine specimens subjected to the face bend test failed prior to
obtaining the required bend. The results organized by specimen number can be
seen in Table 5.5. Selected photographs of a successful and an un-successful
63


specimen can be seen in Figures 5.7 and 5.8. Additional photographs can be
seen in Appendix B. The results of the welded specimen bend tests are
discussed in Chapter VI. Reference Table 4.2 for associated groove angle and
preheat.
Table 5.5 Welded Specimen Bend Test Results
Specimen # Root Bend Test Result Face Bend Test Result
M5-1 Pass Fail
M5-2 Pass Fail
M5-3 Pass Fail
M5-4 Fail Fail
M5-5 Pass Fail
M5-6 Pass Fail
M5-7 Pass Fail
M5-8 Pass Fail
M5-9 Pass Fail
64


Figure 5.8: Welded Specimen Failed Face Bend Test
(Specimen M5-3)
65


CHAPTER VI
DISCUSSION FROM TEST RESULTS
Base Metal Testing Discussion
The tension tests of the base metal specimens taken from Minnesota
Bridge 5721 demonstrated good strength and ductility. The average yield
strength of 36.3 ksi and average ultimate tensile strength of 50.6 ksi (reference
Table 5.1) are within the range of expected values based on comparison with all
reviewed published data on the physical properties of wrought iron (reference
Chapter II). Verifying the strength of this historic wrought iron is crucial to the
ability of engineers to confidently recommend repairs for the metal. If repair
methods that develop the full strength of the original material are used, engineers
can be confident that the age of the material will typically not have an affect on
the tensile strength of the material. Although the values of yield strength and
ultimate strength are within the expected range, it is significant to note that this
wrought iron manufactured approximately 130 years ago, and in constant use for
most of that time, maintained its physical properties so well, and was not
affected by corrosion or fatigue. The low coefficient of variation (CV)
demonstrated in the yield strength and ultimate strength of these base metal
samples displays the fairly constant nature of these properties. This is consistent
with what has been seen by other researchers, such as Gordon and Knopf
(reference Chapter II) who note that slight chemical variations in wrought iron will
affect ductility of the material much more than they will affect the strength of the
66


material (Gordon & Knopf, 2005). In other words, it is to be expected that the
values of Fyand Fu for multiple wrought iron samples will have little variance.
The variability of wrought iron ductility can be seen in the results of the
base metal tension tests. The CV calculated for the measurement of percent
elongation is more than two times the CV calculated for the values of Fy and Fu.
Additionally, the ductility of the wrought iron tested for this study, is very different
from the ductility reported by Bowman (reference Chapter II). In the Purdue
report, Bowman described the noticeable lack of ductility in their specimens
demonstrated by the lack of any visible deformation or necking of the specimen
prior to tensile failure. As can be seen in various photographs (reference Figure
6.1 and Appendix B) each specimen tested for this study went through a visible
necking stage (plastic deformation) prior to failure. Beyond the visible differences
in ductility between the specimens tested here and those tested by Bowman in
the Purdue study, the recorded percent elongation of each study is significantly
different. The average elongation of 11.7% recorded by Bowman is nearly 2.5
times less than the average elongation calculated by the testing performed in this
study (28.9%, reference Table 5.1). Although no chemical analysis was
performed on the wrought iron tested by this study, it can be deduced from the
high ductility of the specimens, that this wrought iron has a relatively low
phosphorus content. As stated in Chapter II of this thesis, an increase in the
phosphorus content of wrought iron will decrease ductility. This is further
demonstrated by the lack of ductility in the specimens of the Purdue study, which
did undergo chemical analysis and were shown to have a phosphorus content
67


between 0.25 and 0.36 percent by weight (Bowman, 2004). This phosphorus
content is nearly three times larger than that shown in the typical wrought iron
chemical analysis of Table 2.1.
The results of the wrought iron base metal specimen tension tests
demonstrate consistent yield strength and ultimate strength, as well as
demonstrating high ductility. When compared to published data provided by
previous wrought iron testing, the values determined from the base metal tension
tests performed in this study, support the conclusion that chemical variations in
wrought iron can have significant impact on the ductility of the material, while not
greatly affecting the yield strength or ultimate tensile strength.
Figure 6.1: Necking Base Metal Tension Test
68


Spark Test Discussion.
The spark test was performed as an inexpensive, efficient method of
verifying the low carbon content, and wrought iron nature of the historic metal
which was being tested. Reference Figure 5.3 for a photograph of the spark test.
The diagram in Figure 6.2 is a sketched depiction of the expected spark patterns
for various types of ferrous metals. It can be seen from items B D in the
diagram that as carbon content in ferrous metal increases, the amount of
forking or splitting at the end of the sparks will increase. Therefore, the low
carbon content of wrought iron should produce fairly straight line sparks, and no
forking at the end of the sparks, as shown in item A of Figure 6.2.
Figure 6.2: Spark Test Diagram (Oberg & Jones, 1918)
A: Wrought Iron, B: Mild Steel, C: Steel w/ 0.5% 0.85% carbon, D: High-carbon
steel, E: High-speed steel, F: Manganese steel, G: Mushet Steel, H: Special
magnet steel
69


The spark test performed on the wrought iron samples from Minnesota
Bridge 5721 clearly demonstrate the spark pattern expected for historic wrought
iron. Based on this test, and the observed silica slag, it can be confidently stated
that the material used for this study is wrought iron.
Welded Specimen Tension Tests Discussion
As described in Chapter V, welded specimens subjected to tension tests
displayed deformation and necking very similar to the un-welded base metal
specimens. Additionally, many of the welded specimens displayed this necking
on both sides of the weld, prior to failure occurring on one side of the weld. The
average yield strength of the welded specimens was 32.2 ksi and the average
ultimate strength was 49.8 ksi. Comparing these values to the averages obtained
from the base metal samples (Fy = 36.3 ksi, Fu = 50.6 ksi), a slight decrease
can be noted (11% for the yield strength and 1.6% for ultimate strength).
However, this slight decrease does not mean that the welding of the wrought iron
decreased the yield or ultimate strength. The decrease in strength was most
likely a result of slight differences in the properties of the wrought iron, as
material from two different eye-bars was used for this process. The consistent
failure location outside the heat affected zone demonstrates that the strength of
the final welded specimen was not negatively affected by the welding process.
The ductility of the welded specimens must also be compared with the
ductility shown by the base metal specimens. The results in Tables 5.1 and 5.2
show that the average percent elongation calculated from the base metal
70


specimens and welded specimens were 28.9% and 25.4%, respectively. Based
on these values, a 12% decrease in the average values of percent elongation
was seen. This decrease in ductility between the welded and non-welded
specimens should was expected because the weld metal did not yield, and that
portion of the sample was not subject to any deformation. The fact that the
average percent elongation of the welded wrought iron specimens was over 25%
leads to the conclusion that the double bevel groove weld which was performed
did not have a significant impact on the ductility of the metal.
In addition to the overall results of all welded specimens, Tables 5.3 and
5.4 display the average values for Fy, Fu and percent elongation, organized by
weld parameter. Table 5.3 displays the average results of each groove angle
used, and Table 5.4 the average results of each preheat temperature used. The
values displayed in these tables make it clear that there is no measurable
difference in the strength or ductility of the sample welded with 30, 45 or 60
degree bevel angles, or of those in which preheat temperatures were used. It
appears that the specimens welded with a 600 F preheat have a slightly lower
Fy and Fu while showing slightly greater ductility. However, the small differences
in these values compared to those of the specimens welded with no preheat or
300 F preheat prevent hard conclusions to be drawn regarding the benefit or
detriment of varying preheat temperatures. Both Tables 5.3 and 5.4 show the
coefficient of variation to be extremely small, which leads to the conclusion that
all nine weld types performed (see Table 4.2) are satisfactory for developing the
full strength and ductility of the historic wrought iron.
71


Although no significant difference between the different types of welds can
be seen from the testing results, the welding process must be taken into account
when recommending a new welding procedure. Because there are so many
factors that can adversely affect the weld, it is important that a welding procedure
provide as much opportunity as possible for consistent quality. If the required
welding procedure is easier to perform, the likelihood of weld defects and flaws
will be greatly decreased. For the welds performed in this study, it was
determined that a 30 degree groove angle is extremely narrow for a standard
electrode, and obtaining full penetration on the root weld pass is difficult.
According to the individuals performing the welds, even with a 45 degree groove
angle full penetration and manipulation of the welding rod was difficult, and the
60 degree groove angle provided the best opportunity for a successful weld.
Regarding preheat temperature it was determined by the welders that using a
preheat temperature of 300 F provided the best welding environment, as
applying a 600 F preheat seemed to create too much preheat in the metal and
caused the welds to be more difficult to perform. See Chapter VII for suggestions
for future research and conclusions based on the research performed here.
Welded Specimen Bend Tests Discussion
The results of the guided bend tests of the welded specimens present an
intriguing challenge to the goal of developing an acceptable welding procedure
for historic structural wrought iron. Of the nine specimens tested with the root
bend test, one specimen failed, and eight passed. However all nine of the
72


specimens tested with the face bend test failed prior to reaching the desired 180
degree bend. Reference Chapter V for a summary of these results. In standard
welding practice and quality control, the failure of a guided bend test indicates a
flaw or defect in the weld. However, it is highly unlikely that poor welding was the
cause of failure for all of the face bend test specimens. The failure of the face
bend tests and the success of the root bend tests is the same phenomenon that
was observed by Chomsrimake in his research (reference Chapter II). The hope
of this study was that performing welds at various preheat temperatures and with
various groove angles would lead to more successful results of the face bend
tests. Each of the nine failed specimens had very similar failure locations. All of
the face bend test failures occurred at the edge of the fusion zone, where the
weld metal and the base metal join. See Figure 5.7 for a photograph of this type
of failure. From the results obtained here, it is clear that the bend test failures
were independent of the preheat temperature and the groove angle.
While the failure of the bend tests does mark the need for continued
research into these wrought iron welding procedures, it is important to note that
the bend tests are required only for welder qualification. The tension tests
performed on the welded wrought iron specimens demonstrate that the welding
procedure described in Chapter IV fully maintains the physical characteristics of
the historic metal.
73


CHAPTER VII
CONCLUSIONS & RECOMMENDATIONS
Conclusions from Research
The wrought iron specimens taken from Minnesota Bridge 5721 provided
important information and aided to the effort of providing acceptable repair
methods for historic structural wrought iron. The series of tests performed on the
base metal and the welded wrought iron specimens yielded valuable data and
provide valuable insight to the ongoing research of the repair of wrought iron
structures. The spark tests performed on the wrought iron samples gave the
researchers confidence that the metal being tested was indeed historic wrought
iron, even without the opportunity to perform detailed chemical analysis. The
base metal tension tests provided crucial data defining the physical properties of
the historic metal, including yield strength, ultimate strength, ductility and
modulus of elasticity. These recorded physical properties were compared with
and verified against values found in previously published data on historic wrought
iron. The base metal tension tests clearly demonstrated the remarkable and long-
lasting strength and ductility of historic structural wrought iron. The longevity of
this structural material is a key reason why many historic wrought iron structures
are still in use today, and why tested and published repair methods are such a
necessity for historic preservation.
While the base metal tests provided necessary data, and valuable
information, the goal of the research described in this thesis was the
74


development of an acceptable welding procedure for historic structural wrought
iron using modern welding techniques. The welding plan used for this research
was created with efficiency and weldability in mind. The use of the double bevel
weld reduced the amount of heat required compared to a single bevel weld, and
the parameters of groove angle and pre-heat were varied with the goal of
determining the most effective welding procedure for the historic wrought iron
metal. The tension tests performed on the welded wrought iron specimens, and
the recorded results, demonstrate that variations in the welding procedures did
not significantly impact any of the measured physical properties of the wrought
iron. As discussed in Chapter VI of this thesis, the data from the tension tests did
not reveal any significant difference in the structural properties of the specimens
welded with various preheat temperatures and groove angles. However, with
weldability and quality in mind, it is recommended that a groove angle of 60
degrees be used with a preheat temperature of 300 F, based on this study.
These parameters will give the welder the best opportunity to perform a high
quality weld. The success of the tension tests of the welded wrought iron
specimens clearly shows that the welding procedures specified in this thesis
(reference Chapter IV) will develop the full strength of the base metal.
Although all but one of the welded specimens passed the root bend tests,
unfortunately, all nine of the welded specimens failed the face bend test
(reference Chapter V). The consistent failure of the face bend tests prevents
these specific welding procedures from being recommended as pre-qualified
welding procedures (reference Chapter III). However, the data collected from all
75


of these tests does provides valuable information for the ongoing effort to
develop a pre-qualified welding procedure for historic structural wrought iron. The
following section of this thesis provides suggestions for future research based on
the findings of the research performed.
Recommendations for Future Research
Welding Methods. Materials
There are many welding methods that have the potential to be used for
the repair of historic wrought iron structures. As discussed in Chapter III, both
GMAW and GTAW methods have characteristics that could enable them to be
used efficiently for the welding of wrought iron. However, there is still research to
be done regarding the use of SMAW welding for wrought iron repair. All the
previously conducted research discussed in this thesis, as well as the tests
performed for this study, utilized E7018 electrodes. It is my hope that in the
future, research will be conducted using multiple electrode types. Better bend
test results might be obtained using electrodes with different chemical
composition and material properties.
Besides research with multiple types of electrode, research should be
performed to investigate the effect of post weld heat treatment (PWHT) on
welded wrought iron. PWHT is used in some modern welding applications in
order to reduce the potential for hydrogen induced cracking, as well as to relieve
residual stresses that are present in the metal. Post weld heat treatment of
76


welded wrought iron may relieve some stresses present in the heat affected zone
and allow the bend tests to be successful.
Joint Types
Although only the butt joint with a double bevel weld was used for this
research, it is my hope that further studies will be performed using many other
joint types, in order to provide engineers and contractors with many available
options for the welded repair of historic wrought iron structures. The joint types
that are most commonly used for modern welding in the construction industry,
such as lap joints, and T-joints (reference Figure 3.1), should especially be
investigated for their potential to be used in pre-qualified welding procedures for
historic wrought iron. Currently, wrought iron welding research has only been
performed using V-groove bevel welds, but hopefully, future research will include
wrought iron welding procedures involving fillet welds on various joint types.
Wrought Iron Material Investigation
In addition to research investigating welding procedures for historic
structural wrought iron, there is a continued need for research into the chemical
and physical properties of various historic wrought iron structures. Detailed
chemical and physical analysis should be performed on samples of wrought iron
from bridges that are still in use. This would greatly increase the confidence with
which engineers could design retrofit and renovation plans for these bridges. The
varying nature of wrought iron used in historic structures throughout the world
77


(reference Chapter II) creates the need for detailed study of the wrought iron
material in a structure prior to repairs being made. It is my hope that full chemical
analysis, such as that shown in Table 2.2, will eventually be performed on the
wrought iron material taken from Minnesota Bridge 5721.
The research on the ability of double bevel V-groove welds to repair
historic wrought iron taken from Minnesota Bridge 5721 proved to be a valuable
step towards the goal of developing a pre-qualified welding procedure for use in
the repair of historic wrought iron structures.
78


REFERENCES
Aston, J. & Story, E. (1939). Wrought Iron: Its Manufacture, Characteristics and
Applications (2nd Edition), A.M. Byers & CO, Pittsburgh, PA
AWS (1991). Specification for Carbon Steel Electrodes for Shielded Metal Arc
Welding: An American National Standard, American Welding Society, Miami, FL
AWS (2001). Standard Welding Terms and Definitions, American Welding
Society, Miami, FL
Bowman, M. & Piskorowski, A. (2004). Evaluation and Repair of Wrought Iron
and Steel Structures in Indiana. Joint Transportation Research Program, West
Lafayette, IN
Chomsrimake, P. (2012) Arc Welding Procedure for Repairing Wrought Iron in
Historic Bridges. M.S. thesis, University of Colorado Denver, Denver, CO
Gordon, R. & Knopf, R. (2005). Evaluation of Wrought Iron for Continued
Service in Historic Bridges, Journal of Materials in Civil Engineering, 17(4),
393-399
Johnson, R. & Olson, S. (2012). Collaboration and Innovation Lead to 3rd
Service Life for an Iconic Iron Bridge. Structure Magazine, 19(10), 29-31
Mills, A. & Hayward, H. (1922). The Manufacture of Wrought Iron. Materials of
Construction: Their Manufacture and Properties (2nd Edition), John Wiley &
Sons, New York, NY, 25-29
Oberg, E. & Jones, F. (1918), Iron and Steel (1st Ed), The Industrial Press, New
York, NY
79


Force (lbs)
Appendix A
Tension Test Result Graphs
0 0.5 1.0 1.5 2.0
Displacement (in)
Figure A.1: Specimen B-M3-2 Force vs. Displacement Plot
80


Force (lb)
81


Force (lb)
40000
36000
32000
28000
24000
20000
16000
12000
8000
4000
0
0 0.50 1.00 1.50
Displacement (in)
Figure A.3: Specimen B-M1-2 Force vs. Displacement Plot
82


Force (lb)
Figure A.4: Specimen B-M3-3 Force vs. Displacement Plot
83


Force (lbs)
84


Force (lbs)
85


Force (lbs)
0 0.5 1 1.5
Displacement (in)
Figure A.7: Specimen M1-2 Force vs. Displacement Plot
86


Force (bs)
87


Force (lbs)
50000
45000
40000
35000
30000
25000
20000
15000
10000
5000
0
Displacement (in)
Figure A.9: Specimen M1-4 Force vs. Displacement Plot
88


Full Text

PAGE 1

INVESTIGATION OF DOUBLE BEVEL FULL PENETRATION GROOVE WELDS FOR THE REPAIR OF HISTORIC STRUCTURAL WROUGHT IRON by JOEL D. WATTERS B.S., University of Southern California, 2007 A thesis submitted to the Faculty of the Graduate School University of Colorado in partial fulllment of the requirements for the degree of Masters of Science Civil Engineering 2013

PAGE 2

This thesis for the Master of Science Degree by Joel D. Watters has been approved for the Department of Civil Engineering by Frederick Rutz, Chair Kevin Rens Cheng Yu Li November 12, 2013 ii

PAGE 3

Watters, Joel D. (M.S., Civil Engineering) Investigation of Double Bevel Full Penetration Groove Welds for the Repair of Historic Structural Wrought Iron Thesis directed by Assistant Professor Frederick R. Rutz ABSTRACT Structural wrought iron was used extensively in the second half of the 19th Century to construct bridges and buildings throughout the United States. Many of these wrought iron structures are still in use today, and are in need of repair and renovation. While many methods are used for the repair of these structures, there is no published, pre-qualied welding procedure using modern welding technology for the repair of historic structural wrought iron. This thesis investigates the ability of a specic welding procedure, using modern welding technology, to repair structural wrought iron samples which were removed from a historic bridge. Additionally, the physical properties of this historic structural wrought iron were tested and documented, in an effort to increase the data available to engineers who are tasked to work with this unique structural metal. The form and content of this abstract are approved. I recommend it's publication Approved: Frederick Rutz iii

PAGE 4

ACKNOWLEDGEMENTS This thesis and the research conducted for its completion were made possible by the tireless dedication of Dr. Frederick R. Rutz. Through his efforts, large amounts of historic structural wrought iron were made available for this important research, and his passion and dedication have enabled continued research in the hopes of developing suitable welding procedures for this unique structural metal. Thank you to the faculty and welding students of Western Colorado Community College of Grand Junction, CO for their support and investment of time into this research. Their excellent work and efforts made the completion of this thesis possible. Thank you to Ryan Thomas of Emily Grifth Opportunity School in Denver, CO for his invaluable training in the shielded metal arc welding process, and for his knowledgable advice regarding the research conducted for this thesis. Thank you to Tom Thuis and Jac Corless of the Civil Engineering lab, and machine shop at University of Colorado Denver. They provided extensive assistance in the preparation of wrought iron test specimens, and in the operation of the testing equipment used. iv

PAGE 5

TABLE OF CONTENTS CHAPTER I. INTRODUCTION !!!!!!!!! 1 Necessity of Research !!!!!!! 1 Scope of Research !!!!!!!! 2 Goals for Research !!!!!!!! 3 II. HISTORY & LITERATURE REVIEW !!!!!! 5 History of Wrought Iron !!!!!!! 5 !! Manufacture of Wrought Iron !!!!! 5 !! Structural Applications of Wrought Iron !! !! 7 !! Properties of Wrought Iron !!!! !! 8 !! Minnesota Bridge #5721 !!!!!! 15 Repair & Preservation of Structural Wrought Iron !!! 17 !! Testing of Historic Structural Wrought Iron !!! 18 !! Welding Repair of Structural Wrought Iron !!! 19 III. WELDING BACKGROUND !!!!!!! 25 Welding Methods !!!!!!!! 25 !! Shielded Metal Arc Welding !!!!! 25 !! Gas Metal Arc Welding !!!!!! 29 !! Gas Tungsten Arc Welding !!!!!! 29 Standard Joint Types !!!!!!! 30 Weld Quality, Flaws & Testing !!!!!! 32 Welding Procedure Qualication !!!!!! 35 v

PAGE 6

IV. WROUGHT IRON TESTING TEST DESIGN & CONFIGURATION 37 Base Metal Tension Tests Design & Conguration !!! 37 !! Test with 220k MTS !!!!!!! 37 !! Test with 20k MTS !!!!!!! 44 Base Metal Spark Test Design & Conguration !!! 46 Welded Specimen Tension Tests Design & Conguration !! 46 Welded Specimen Bend Tests Design & Conguration !! 50 V. WROUGHT IRON TESTING RESULTS !!!!! 56 Base Metal Testing Results !!!!!! 56 Spark Test Results !!!!!!!! 60 Welded Specimen Tension Tests Results !!!! 61 Welded Specimen Bend Tests Results !!!!! 65 VI. DISCUSSION FROM TEST RESULTS !!!!! 68 Base Metal Testing Discussion !!!!!! 68 Spark Test Discussion !!!!!!! 71 Welded Specimen Tension Tests Discussion !!!! 72 Welded Specimen Bend Tests Discussion !!!! 74 VII. CONCLUSIONS & RECOMMENDATIONS !!!!! 76 Conclusions from Research !!!!!! 76 Recommendations for Future Research !!!!! 78 !! Welding Methods, Materials !!!!! 78 !! Joint Types !!!!!!!! 79 !! Wrought Iron Material Investigation !!!! 79 vi

PAGE 7

REFERENCES !!!!!!!!! 81 APPENDIX !!!!!!!!!! A. Tension Test Plots !!!!!!! 82 B. Photographs Tension Tests !!!!!! 99 C. Photographs Bend Tests !!!!! 102 vii

PAGE 8

TABLES TABLE 2.1 Typical Wrought Iron Chemical Analysis !!!!! 14 2.2 Early 20th Century Steel Chemical Analysis !!!! 15 2.3 Results of Tensile Tests by Gordon & Knopf !!!! 19 2.4 Results of Tensile Tests by Chomsrimake !!!! 21 2.5 Results of Tensile Tests by Bowman & Piskorowski !!! 23 3.1 Carbon Steel Covered Arc Welding Electrodes !!!! 28 3.2 Applicable Welds for Basic Joint Types !!!!! 32 4.1 Weld Parameters for Welded Tension Tests !!!! 48 4.2 Weld Parameters for Welded Bend Tests !!!! 51 5.1 Base Metal Tension Test Results !!!!!! 57 5.2 Welded Specimen Tension Test Results !!!!! 63 5.3 Welded Specimen Tension Test Results By Groove Angle !! 63 5.4 Welded Specimen Tension Test Results By Preheat Temperature 64 5.5 Welded Specimen Bend Test Results !!!!! 66 viii

PAGE 9

FIGURES FIGURE 2.1 Wrought Iron Photomicrograph !!!!! 9 2.2 Ruptured Wrought Iron Specimen !!!!! 10 2.3 Wrought Iron Physical Properties !!!!!! 12 2.4 Minnesota Bridge 5721 Stillwater, MN !!!!! 16 2.5 Single V-Groove Weld !!!!!!! 20 3.1 Standard Joint Types !!!!!!! 31 4.1 220k MTS Testing Frame !!!!!!! 38 4.2 20k MTS Testing Frame !!!!!!! 38 4.3 Design of Steel Clamp Bar for Tension Test !!!! 40 4.4 Design of Wrought Iron Tension Test Specimen !!! 42 4.5 Base Metal Tension Test Setup !!!!!! 43 4.6 Wrought Iron Specimen for 20k MTS with Extensiometer !! 45 4.7 Eye-Bar M1 Segmentation Diagram !!!!! 47 4.8 Double Bevel Full Penetration Groove Weld !!!! 49 4.9 Lincoln Electric Precision TIG 185 !!!!! 49 4.10 Eye-Bar M5 Segmentation Diagram !!!!! 50 4.11 Preparation of Bend Test Specimen !!!!! 52 4.12 Prepared Bend Test Specimens !!!!!! 53 4.13 Guided Bend Test Schematic !!!!!! 54 4.14 Watts W-50 Tensile & Bend Weld Tester !!!!! 55 5.1 B-M1-1 Force vs. Displacement Plot !!!!! 58 ix

PAGE 10

5.2 B-M1-2: Stress vs. Strain Plot !!!!!! 59 5.3 Spark Test Photograph Wrought Iron from MN Bridge 5721 !! 60 5.4 Spark Test Photograph Mild Steel !!!!! 61 5.5 Welded Specimen Tension Test Photograph 1 !!! 64 5.6 Welded Specimen Tension Test Photograph 2 !!! 65 5.7 Welded Specimen Successful Root Bend Test !!! 67 5.8 Welded Specimen Failed Face Bend Test !!!! 67 6.1 Visible Ductility Base Metal Tension Test !!!! 70 6.2 Spark Test Diagram !!!!!!! 71 A.1 Specimen B-M3-2 Force vs. Displacement Plot !!! 82 A.2 Specimen B-M1-1 Force vs. Displacement Plot !!! 83 A.3 Specimen B-M1-2 Force vs. Displacement Plot !!! 84 A.4 Specimen B-M3-3 Force vs. Displacement Plot !!! 85 A.5 Specimen B-M3-4 Force vs. Displacement Plot !!! 86 A.6 Specimen B-M1-1 Force vs. Displacement Plot !!! 87 A.7 Specimen B-M1-2 Force vs. Displacement Plot !!! 88 A.8 Specimen B-M1-3 Force vs. Displacement Plot !!! 89 A.9 Specimen B-M1-4 Force vs. Displacement Plot !!! 90 A.10 Specimen B-M1-5 Force vs. Displacement Plot !!! 91 A.11 Specimen B-M1-6 Force vs. Displacement Plot !!! 92 A.12 Specimen B-M1-7 Force vs. Displacement Plot !!! 93 A.13 Specimen B-M1-8 Force vs. Displacement Plot !!! 94 A.14 Specimen B-M1-9 Force vs. Displacement Plot !!! 95 x

PAGE 11

A.15 Specimen B-M5-10 Force vs. Displacement Plot !!! 96 A.16 Specimen B-M5-11 Force vs. Displacement Plot !!! 97 A.17 Specimen B-M5-12 Force vs. Displacement Plot !!! 98 B.1 Tension Test with 220k MTS !!!!!! 99 B.2 Tension Test Specimen !!!!!!! 99 B.3 Tension Test with 220k MTS !!!!!! 100 B.4 Tension Test with 20k MTS !!!!!! 101 C.1 Prepared Bend Test Specimens !!!!!! 102 C.2 Bend Test Operation !!!!!!! 102 C.3 Failed Face Bend Test !!!!!!! 103 C.4 Successful Root Bend Test !!!!!! 103 xi

PAGE 12

ABBREVIATIONS Avg. Average AWS American Welding Society CV Coefcient of Variation (Std. Dev. / Mean) GMAW Gas Metal Arc Welding GTAW Gas Tungsten Arc Welding k "kip," a unit of force, (1k = 1000 lbs) SMAW Shielded Metal Arc Welding Std. Dev. Standard Deviation WPS Welding Procedure Specication WPQR Welding Procedure Qualication Record xii

PAGE 13

CHAPTER I INTRODUCTION Necessity of Research In recent years the deteriorating state of infrastructure in the United States has become increasingly clear. As roads, bridges and other infrastructure components that were built over 100 years ago begin to require more and more maintenance, it is vital that the means for repair and rehabilitation of existing structures are in place. Due to limited budgets for infrastructure spending, it cannot be expected that deteriorating infrastructure components can always be replaced with new ones. Although much of the infrastructure in disrepair today is built of concrete and steel, there are still a large number of structures, mostly bridges, in the United States that incorporate structural wrought iron as a signicant part of their structural system. These bridges, most of which were built over 100 years ago carry rail, vehicle and/ or pedestrian trafc. Although little construction has utilized structural wrought iron in the past century, it is important for the engineering and construction community to have access to a broad range of repair methods for the rehabilitation of existing wrought iron bridges. Modern welding techniques have the potential to give engineers and preservation contractors many ways to repair historic structural wrought iron. Although welding has been performed on historic structural wrought iron, there is a severe shortage of documented testing and approval of welding procedures for this unique structural metal. Because most historic wrought iron structures are 1

PAGE 14

bridges, their repair falls under the oversight of the local or state department of transportation. For repairs that will affect the structural system of an existing bridge, these governing bodies typically require that any welding performed conforms to a pre-qualied welding procedure. The majority of pre-qualied welding procedures published by the American Welding Society (AWS) apply to steel and other modern structural metals, and approved repair methods using modern welding techniques on historic structural wrought iron are unavailable to engineers and contractors. For this reason, the research and development of welding procedures for historic structural wrought iron is vital to the preservation of existing infrastructure. Scope of Research Although the possibilities of welding historic wrought iron are very extensive, the scope of research presented herein is limited due to time and resources available within the Master's Thesis program. As discussed in more detail in Chapter III, many welding methods may have potential use for the repair of historic wrought iron. However, the research conducted for this thesis was limited to a specic joint type and a specic welding rod. Due to the iterative nature of this research, it was paramount that as many variables as possible were kept constant, ideally only changing one variable at a time. For this project, an E7018 welding rod was used, and the weld type was a double-bevel full penetration groove weld on a butt joint (reference Chapter III). The parameters that were iteratively adjusted included the pre-heat temperature of the joint, and 2

PAGE 15

the groove angle of the weld joint. For a full description of the welding procedures which were tested, refer to Chapter IV. In addition to limiting the welding methods and joint types used, the tests performed on the base metal and the welded specimens were limited to tension tests, root bend tests, face bend tests, and spark tests. For more information regarding the design and conguration, and the results of these tests, refer to Chapters IV and V. Goals for Research The goal for this project is to develop a satisfactory welding procedure to be used in the repair of historic structural wrought iron. Although simply stated, to prove that a specic welding procedure will provide the required structural integrity, it must be shown that the weld fully develops the strength of the base metal. Additionally, it must be shown that the welding process does not signicantly reduce the ductility of the base metal. For an existing wrought iron structural member subject to exural stresses, reduction in ductility could lead to premature, and possible brittle failure of the member. The tension tests described in Chapter IV of this thesis aim to demonstrate that the weld performed will not reduce the strength or ductility of the wrought iron specimen, and will not change the failure mechanism of the specimen. The bend tests are designed to properly qualify the welding procedure and to verify proper welding technique. In short, the goal of the tests performed was to show that the full penetration weld of two wrought iron pieces will allow the nal welded specimen to behave the same or in 3

PAGE 16

an improved manner when compared to the behavior of an un-welded specimen of the same material. 4

PAGE 17

CHAPTER II HISTORY & LITERATURE REVIEW History of Wrought Iron Manufacture of Wrought Iron Before steel became the primary structural metal for the construction industry, wrought iron had, for the previous 50 years, been used to construct buildings and bridges. Although all three of these products are made from iron, and to the naked eye may look very similar, the processes by which they are and were manufactured give them very different properties. As steel became better understood and more efciently produced in the early 1900s, wrought iron and cast iron were mostly phased out as structural metals. During the last half of the 19th Century, however, wrought iron was a common structural material, and especially during the period of 1850-1880, many bridges and buildings were constructed using wrought iron as a main structural component. The invention of the grooved rolling process for wrought iron was the critical development that allowed wrought iron to be produced in large quantities, and used in the construction of large bridges and buildings. Before the invention of grooved rolling mills, wrought iron and other metals, could only be rolled into at sheets, which limited the nal shapes that could be produced. The typical manufacture process of wrought iron in the 19th Century consisted of, "the melting of the pig iron in the hearth of a reverbatory furnace which is lined with iron oxides, resulting in the elimination of most of the carbon, 5

PAGE 18

silicon, manganese, phosphorus, and sulphur present in the charge by oxidation" (Mills & Hayward, 1922). The melted pig iron was then rolled in to bars called muck bars which were subsequently bundled together and rolled again to produce merchant bars. According to Adelbert Mills and Harrison Hayward in their publication entitled Materials of Construction: Their Manufacture and Properties, This process results in the production of very pure iron mixed with from 1% to 3% of slag, which the rolling process has caused to assume the form of greatly elongated particles in the direction of rolling. This circumstance accounts for the characteristic brous structure of wrought iron" (Mills & Hayward, 1922). These merchant bars were also referred to as single-rened iron, describing the single re-rolling of the original muck bars. In many instances, the merchant bars would be piled together and rolled again, producing double-rened bars. This repeated rolling of the iron produced, "further elongation of the strands of slag in the direction of rolling, thereby rendering the iron still more brous in its structure" (Mills & Hayward, 1922). As discussed in the "Properties of Wrought Iron" section of this Thesis, the brous structure created by the iron silicate is a unique and important feature of historic wrought iron. Structural Applications of Wrought Iron The continued improvement of the rolling process through the rst half of the 19th century enabled wrought iron to be used in increasingly wider applications. Enhancements to the rolling process allowed wrought iron to be 6

PAGE 19

rolled in various structural shapes. Additionally, during the middle of the 19th Century, as the railroad industry was booming across the United States, many bridges built of cast iron had collapsed, so builders began to turn to wrought iron as a more ductile and reliable structural metal. During this period, while there was a continual increase in the production and use of wrought iron, the rst rolled wrought iron structural shapes were produced. While Ferdinand Zores of Paris is apparently the rst person to have successfully produced rolled wrought iron I-beams, Peter Cooper was the rst to roll structural shapes in the United States, at his Trenton Iron Works in New Jersey. Although these processes enabled the rolling of wrought iron structural shapes, the production of these shapes was still limited by the size of iron ingot that could be processed in the furnaces of that time. Due to this limitation, large wrought iron beams had to be created by the riveting together of smaller, more manageable sections. Several buildings in the United States which used wrought iron beams were constructed during the second half of the 19th Century. It was not until the development of the Bessemer process for the rening of steel, as well as the invention of elevators that buildings were able to be built taller than six stories and be constructed of a steel skeleton, without the use of masonry, allowing for more window space. These developments led to the beginning of the skyscraper era in the rst half of the 20th Century. 7

PAGE 20

Properties of Wrought Iron In the 19th Century, prior to the creation of governing bodies and organizations with the authority to regulate the production of structural metals, wrought iron, and other structural materials were commonly manufactured with varying properties. In many cases the wrought iron produced by one renery would not have the same properties as that produced by another renery. Although lacking the strict standardization of modern structural metals, there are characteristics that are common to historic wrought iron, regardless of where the metal was produced. The rst, and perhaps most distinctive, characteristic of wrought iron is the slag that is present in the metal. The denition of wrought iron according to the American Society for Testing and Materials (ASTM) from 1930, states that wrought iron is, "a ferrous material, aggregated from a solidifying mass of pasty particles of highly rened metallic iron, with which, without subsequent fusion, is incorporated a minutely and uniformly distributed quantity of slag" (Aston & Story, 1939). This quote is taken from a publication by James Aston and Edward B. Story entitled Wrought Iron: It's Manufacture, Characteristics and Applications. In this publication, Aston and Story go on to describe this slag that is unique to historic wrought iron. They state that the, "slag content in wrought iron may vary from about 1% to 3% or more, by weight," and "in well made wrought iron, there may be 250,000 or more of these glass-like slag bers to each cross-sectional square inch" (Aston & Story, 1939). Figure 2.1 shows a photo with 100X magnication, clearly displaying the brous slag embedded in the base material. 8

PAGE 21

As can be seen in Figure 2.2, even in an un-magnied photograph of a ruptured specimen, the brous nature of the wrought iron is visible. Figure 2.1: Wrought Iron Photomicrograph (Aston & Story, 1939) Figure 2.2: Ruptured Wrought Iron Specimen 9

PAGE 22

! While steel has only a crystalline structure, wrought iron has, in addition to its crystalline iron structure, a brous structure which is essential to the ductility of the material, as well as it's high resistance to fatigue and corrosion. However, due to the single direction distribution of the slag bers, caused by the rolling process, wrought iron has a much higher tensile strength in the direction of rolling than in the direction orthogonal to rolling. In other words, wrought iron has a higher ultimate tensile strength (F u ) and yield strength (F y ) in the longitudinal direction. Because the physical properties of wrought iron "are largely those of pure iron," the quantity and distribution of the brous slag has a great effect on the physical properties of the wrought iron sample (Aston & Story, 1939). Figure 2.3 shows a compiled list of the strength of various wrought iron members. These values were published in the catalogs of rolling mills during the last 30 years of the 19th Century. In addition to its high ductility compared to that of cast iron, wrought iron was used extensively in the 19th Century, because it showed high resistance to fatigue. According to Aston and Story, Wrought iron is known to be relatively insensitive to notch effects and unusually resistant to over stress. These desirable properties are attributed primarily to high ductility, and of particular importance, to the presence of the slag bers which confer on the metal a tough, brous structure somewhat analogous to that of a stranded wire cable (Aston & Story, 1939). They go on to state that, "the slag bers apparently serve to minimize stress concentration and deect the path of slip planes that develop in a metal under the inuence of conditions which ordinarily would result in fatigue failure" (Aston & Story, 1939). These desirable properties enabled wrought iron to be the 10

PAGE 23

dominant structural metal for most of the 19th Century, until steel was able to be produced with more reliable and desirable properties and in mass quantities. The list shown in Figure 2.3 does not list any values for ductility or elongation percent of the material. This is because through most of the 19th Century, engineers were typically given only values of the yield and ultimate strength of a material. In an article published in 2005 in the Journal of Materials in Civil Engineering entitled Evaluation of Wrought Iron for Continued Service in Historic Bridges Robert Gordon and Robert Knopf state that, "One of the earliest quantitative ductility specications was published in the 1873 account of the construction of the International Bridge over the Niagara River at Buffalo, NY" (Gordon & Knopf, 2005). Although historic wrought iron had a reputation for high ductility and fatigue resistance, small variations in the chemical composition of wrought iron, while not greatly affecting the ultimate strength, were known to have signicant effects on the metal's ductility. Similar to the test results presented in this thesis (see Chapters IV through VI) Gordon and Knopf discuss the wide range of ductility found in tested wrought iron specimens from the 19th Century, "As the test data show, a wide range of ductilities can be attained in wrought iron, either intentionally, or in the hands of less skilled makers, by accident" (Gordon & Knopf, 2005). The information presented by Gordon and Knopf describe methods for estimating ductility of existing wrought iron structural members without destructive testing. 11

PAGE 24

Figure 2.3: Wrought Iron Physical Properties (Aston & Story, 1939) In addition to the physical properties of wrought iron, the chemical properties of this unique structural metal must be understood in the context of the inherent slag distribution. For chemical analysis of historic wrought iron, the twocomponent nature of the metal must be taken into account. A thorough analysis of wrought iron should provide the distribution of chemicals in each component, both the slag and the iron. A typical wrought iron specimen will likely contain small amounts of the following metalloids: carbon, phosphorus, manganese, silicon and sulfur. Table 2.1 provides the chemical distribution of a typical wrought iron sample. For comparison, Table 2.2, a chemical analysis from a typical early 12

PAGE 25

20th Century mild steel, has been provided. As can be seen from these tables, the carbon content of wrought iron (0.02% in sample shown here) is signicantly lower than that of steel, although in some cases, "good wrought iron may have a carbon content of 0.08% or 0.10%" (Aston & Story, 1939). Higher amounts of carbon in a wrought iron sample may create a higher strength wrought iron, but will also signicantly decrease the ductility of the material. A higher carbon content than what is expected in wrought iron, "may be an indication of imperfect or incomplete rening, or may awaken suspicion that steel scrap has been used in bushelling or piling [added during the forging of the iron]" (Aston & Story, 1939). Another signicant difference between historic wrought iron and early steel is the manganese content. In the words of Aston and Story, "The virtual absence of manganese in wrought iron and its almost universal presence in steel has resulted in the manganese determination being used as a means of identication and differentiation" (Aston & Story, 1939). The chemical phosphorus has a signicant impact on the ductility of the wrought iron; wrought iron with a higher phosphorus content will display reduced ductility. Seemingly slight variations in the chemical content of wrought iron, incurred during the manufacturing stage, have the ability to produce metal with greatly varied physical properties. For this reason, where possible, it is important to perform a detailed chemical analysis of the historic wrought iron sample in question. However, due to nancial and scheduling limitations, a chemical analysis was not performed on the material that has been tested for this study. A simple "spark test" was performed to verify the wrought iron nature of the 13

PAGE 26

material, as well as visual inspection. See Chapters IV VI for the conguration, results and discussion of the spark test. Even without a chemical analysis, conclusions can be drawn regarding the chemical content of the tested material based on how the material performs under physical testing. See Chapters VI and VII for this discussion and analysis. Table 2.1: Typical Wrought Iron Chemical Analysis (Adapted from Aston & Story, 1939) Combined Analysis (%) Separate Analysis (%) Separate Analysis (%) Base Metal Slag Carbon 0.02 0.02 Manganese 0.03 0.01 0.02 Phosphorus 0.12 0.1 0.02 Sulphur 0.02 0.02 Silicon 0.15 0.01 0.14 Slag by Weight 3 Table 2.2: Early 20th Century Mild Steel Chemical Analysis (Adapted from Aston & Story, 1939) Metalloid Content (%) Carbon 0.1 Manganese 0.5 Phosphorus 0.04 Sulphur 0.05 Silicon 0.1 14

PAGE 27

Table 2.3: Results of Tensile Tests by Gordon and Knopf (Adapted from Gordon & Knopf, 2005) Y u = Upper Yield Stress, Y l = Lower Yield Stress, T = Ultimate Tensile Strength r = Area Reduction, e = Elongation Y u (ksi) Y l (ksi) T (ksi) r (%) e (%) 39.9 39.9 55.7 34 28 35.0 32.8 49.9 35 26 32.1 32.1 48.2 34 26 39.3 38.9 53.1 34 27 46.8 37.7 52.5 33 26 38.9 38.9 54.8 33 27 Average 38.7 36.7 52.4 33.8 26.7 Minnesota Bridge #5721 All the material used for the testing and analysis performed for this thesis was donated by the Minnesota Department of Transportation (MnDOT), and was taken from a bridge referred to as "Minnesota Bridge #5721." This bridge, a 162 ft long Parker truss structure, was built in 1877 in the town of Sauk Centre, MN. For 60 years it, "enabled horses, wagons, buggies and pedestrians to cross the Sauk River" (Johnson, 2012). By the third decade of the 20th Century, the bridge was no longer sufcient for the increased demand caused by the population growth of Sauk Centre and rapidly growing automobile use. Because of this, in 1937, Minnesota Bridge #5721 was "disassembled and in 1937 was re-erected in a wilderness area in northern Minnesota," and for 70 years it "carried logging trucks and other vehicles at this scenic northern Minnesota location" (Johnson, 2012). 15

PAGE 28

! After 70 years of service in it's second location, Minnesota Bridge #5721 was once again deemed to be insufcient for its required use. Both the limited size of the bridge, and it's deteriorating structure made it clear that the bridge could no longer be used for modern highway trafc. In 2009 the bridge was removed from it's second home and stored until 2011 in a MnDOT facility. After a detailed structural study of the bridge, and a feasible rehabilitation plan was completed, the bridge was installed in 2011 into its third and current location near Stillwater, MN (see Figure 2.4). The renovations carried out on the bridge included the replacement of the original deck with a lightweight concrete deck and the replacement of deck stringers and oor end beams (originally wrought iron), with new steel members. Additionally, eight of the 96 original wrought iron eye-bar truss members were replaced with new steel members. The need to replace eight of the original eye-bars was created during the disassembly of the bridge in 2009 when, "the disassembly contractor's hardened tools produced nicks and gouges in the heads of the eye-bars as the pinned connections were taken apart" (Johnson, 2012). 16

PAGE 29

Figure 2.4: Minnesota Bridge 5721 Stillwater, MN (Image used under Creative Commons Attribution-Share Alike 3.0 Unported License) Six of the eight eye-bars removed from the bridge were donated to the University of Colorado Denver in 2012. The test specimens used for this study were taken from two of the eye-bar sections that were removed during the latest rehabilitation and relocation of Minnesota Bridge #5721. Repair & Preservation of Structural Wrought Iron As described in the previous section, since the dawn of the steel era, modern manufacturing and modern construction methods, the need for new structural wrought iron has mostly disappeared. However, because the wrought iron manufactured in the late 19th Century was usually very durable, there are still many structures in use today that rely on wrought iron as signicant portions of their structural systems. It is critical that engineers and contractors tasked with the repair and rehabilitation of these structures be familiar with available and 17

PAGE 30

proven repair methods. In this way, it can be ensured that these historic structures last for many more years, not simply for their historic value but also for their functional role in society. In recent years there has been some research regarding the repair of historic wrought iron using modern construction methods. However, even with the methods that have been studied, there is still a great need for the further study of viable repair options, especially those utilizing modern welding techniques. Chapters IV through VI of this thesis provide a detailed analysis and testing results of the specic welded repair method studied herein. The following sections in this chapter seek to summarize the prominent historic wrought iron repair methods that have recently been studied by other parties. Testing of Historic Structural Wrought Iron Throughout much of the 20th Century, many tests have been performed on structural wrought iron which was manufactured during the 19th Century. These tests include valuable information such as: detailed chemical analysis, evaluation of the elastic modulus, determination of yield strength and ultimate strength as well as measurements of percent elongation and toughness. Examples of some of the available data are displayed in the section "Properties of Wrought Iron," of this thesis. The data shown there was all from material produced in the late 19th and early 20th Centuries. More recent test data can be found in the results published in the article Evaluation of Wrought Iron for Continued Service in Historic Bridges. In this 18

PAGE 31

study, Gordon and Knopf performed physical and chemical testing on wrought iron samples from several historic bridge structures. Similar to the results presented in Chapter V of this thesis, Gordon and Knopf report that, Plastic deformation usually began with a sharp yield point followed by discontinuous yielding that continued until steady work hardening took over, leading to necking down and eventual ductile rupture of the specimen (Gordon, 2005). The results of their tension tests can be seen in Table 2.3. Gordon and Knopf also performed signicant chemical analysis on wrought iron samples. Results from other tests performed on historic structural wrought iron can be found in the following section of this thesis (Tables 2.4 and 2.5). Many other tests have been performed over the last 100 years on historic structural wrought iron from structures built in the 19th Century. Many of these tests show similar results to the ones presented here and the data collected by this study. The discussion of wrought iron testing by other parties in this paper has been limited to include a small number of studies for the purpose of comparison and verifying the legitimacy of the results obtained by this study. Welding Repair of Structural Wrought Iron To the knowledge of this author, only two previously published studies present a thorough investigation of the ability of modern welding techniques to repair historic structural wrought iron. The rst is a study titled Arc Welding Procedure for Repairing Wrought Iron in Historic Bridges, a Masters Thesis by Preeda Chomsrimake at the University of Colorado Denver. 19

PAGE 32

! The study performed by Chomsrimake at the University of Colorado Denver tested the ability of a single bevel groove weld to connect two separate pieces of historic wrought iron. In Chomsrimake's study, 1/8 in. diameter, E7018 rods were used along with a groove made with a 45 degree included angle. These welds were performed using a 1/4 in. by 1 in. steel backing bar, which was removed prior to testing of the specimen (Chomsrimake, 2012). Figure 2.5 shows a cross-section sketch of the weld joint used by Chomsrimake. Figure 2.5: Single V-Groove Weld (Chomsrimake, 2012) Chomsrimake performed tension tests and guided bend tests on the base metal and the welded specimens and tabulated the results. The data from Chomsrimake's tests showed the base metal specimens (no welds) to have an average ultimate strength of 47.5 ksi and a average yield strength of 28.5 ksi. The welded specimens that were subjected to the tension test demonstrated an average ultimate strength of 46.6 ksi and an average yield strength of 28.0 ksi (Chomsrimake, 2012). According to the author, this slight variance in values 20

PAGE 33

between the the base metal specimens and the the welded specimens is not signicant and, "Using SMAW to make V-groove welding in a butt joint of the sample members still maintains the ultimate tension capacity of the historic wrought iron members." (Chomsrimake, 2012). The average elongation reported from these tension tests was 17.1% for the base metal specimens and 9.85% for the welded specimens. These values show a signicant decrease in the ductility of the base metal specimen when compared to the welded specimen and, according to the author, this is likely due to the higher yield strength and stiffness of the weld metal compared to that of the wrought iron base metal. (Chomsrimake, 2012). The summary of the tension test results performed by Chomsrimake can be seen in Table 2.4. Table 2.4: Results of Tensile Tests by Chomsrimake (Adapted from Chomsrimake, 2012) F y (ksi) F U (ksi) Elongation (%) Base Metal Average 28.5 47.5 17.1 Base Metal St. Dev. 1 1.67 2.48 Welded Specimen Average 28 46.6 9.85 Welded Specimen St. Dev. 0.345 0.574 1.66 ! In addition to tension tests of the base metal and welded wrought iron specimens, the 2012 study at the University of Colorado Denver also included guided bend tests of both the root and the face of the V-groove weld. (Reference Chapter IV for description of root and face guided bend tests.) According to Chomsrimake, the root bend tests passed with only very minor cracks visible in the welded zone. However, during the face bend test, the welded wrought iron 21

PAGE 34

specimen "was broken in the base metal near the welded joint" (Chomsrimake, 2012). The other existing study to investigate the ability of modern welding techniques to repair historic structural wrought iron was performed at Purdue University by Mark D. Bowman and Amy M. Piskorowski, entitled Evaluation and Repair of Wrought Iron and Steel Structures in Indiana. In the Purdue study, Bowman and Piskorowski performed tests using several wrought iron repair methods, in an effort to "develop a list of suggested repair techniques for joints and members typical of wrought iron bridge constructions" (Bowman, 2004). Although they performed tests on many repair methods, only their tests on welded wrought iron specimens will be discussed in this thesis. Bowman and Piskorowski used a double bevel V-groove butt joint with an groove angle of 60 degrees and a root opening of 1/16 in. on a 1/2 in. thick wrought iron eye-bar specimen. According to the authors, "Weld passes were alternated on either side [of the joint] to ensure that heat distortion in the piece was minimized" (Bowman, 2004). Regarding the results of the tension tests in general, the authors of the Purdue study state that, "All of the fractures of the historic wrought iron tensile coupon testing were somewhat brittle in nature," and that, "before any of the specimens were about to fail during testing, there was no visible necking or any typical pre-failure behavior that is typically found in structural steel" (Bowman, 2004). Bowman and Piskorowski tested several tensile coupon specimens, ve of which were welded specimens, using the welding process described above. 22

PAGE 35

The summary of their tensile test results is shown in Table 2.5 below. According to the authors, none of the welded tensile tests failed in the region of the weld, and there was "little variation" between the welded and non-welded specimens when comparing both yield strength and ultimate tensile strength. From their analysis and results, the authors of the Purdue study surmise that the welded repair method they used can be "considered a satisfactory weld detail" (Bowman, 2004). Bowman and Piskorowski performed only tension tests on their welded specimens, and as a result do not have any data on ability of the welded repair method to resist bend tests. Nonetheless, they demonstrated that a specic welding technique, using modern welding methods, can be used to sufciently repair a wrought iron member which is subject to tension only. Table 2.5: Results of Tensile Tests by Bowman & Piskorowski (Adapted from Bowman, 2004) E (ksi) F y (ksi) F U (ksi) Elongation (%) Average 27,700 31.6 47.2 11.7 St. Dev. 500 1.28 3.38 4.5 Although the tests and studies performed by other parties, and discussed above, provide valuable information to engineers and those involved in the repair of historic wrought iron structures, it is clear that there is ample need for more studies and analysis, especially those involving the use of modern welding techniques on historic structural wrought iron. 23

PAGE 36

CHAPTER III WELDING BACKGROUND Welding Methods Many modern welding methods have the potential to be useful for the repair of historic structural wrought iron. Three of the most common modern welding techniques will be discussed in the sections below, as they have the highest potential to be used for the successful and economical repair of historic wrought iron. Due to the limited scope of this study, this research investigated only the ability of shielded metal arc welding (SMAW) for wrought iron repair. The methods used, results and discussion of the SMAW process use for the repair of historic structural wrought iron can be found in Chapters IV, V and VI. Shielded Metal Arc Welding The rst, and perhaps most common modern welding technique, especially for in-situ repairs, is the SMAW process. In the shielded metal arc welding process, a ux-covered welding rod acts as the electrode and when placed in proximity to the base metal, the rod forms an electric arc which creates temperatures of 6500-7000 o F. This heat melts both the base metal and the welding rod, and as the rod is melted into the base metal the ux forms a slag (not to be confused with the silica slag present in historic wrought iron) which oats to the top of the weld and hardens as the metal cools. This slag is removed 24

PAGE 37

after each weld pass. With the welding rod acting as both the ller metal and the electrode, SMAW is a simple and easy welding process. The SMAW process can be performed with alternating current (AC) or direct current (DC) power. However, for the research discussed here only DC power was used as DC power is more commonly used in construction applications. Additionally, when DC power is used, with the SMAW process, a welder has the option of using direct current electrode positive (DCEP) or direct current electrode negative (DCEN). These two types of DC welding simply refer to the end of the circuit on which the electrode is placed. For DCEP, the electrons in the circuit ow from the positive pole of the welding machine to the electrode, and the negative pole of the machine is attached to the base metal to complete the circuit. For a DCEN setup, the electrons in the circuit ow from the negative pole of the welding machine to the electrode, and the positive pole of the machine is attached to the base metal to complete the circuit. Typically the decision to use DCEP or DCEN is driven by the welding rod selected for the job, as each rod has certain limitations that govern the welding circuit type on which it can be used. The selection of the type of welding rod is crucial for a successful weld. The choice of welding rod can be based on many factors, including base metal composition, joint type, base metal thickness, and welding environment. For this study, the common and readily available E7018 rod was used. The nomenclature used to designate welding rods for the SMAW process is as follows: the rst two numbers, in this case 70, refer to the minimum required tensile strength of the 25

PAGE 38

ller metal in k/in 2 (ksi). The third number in the rod designation species the recommended welding position for that rod type (at, horizontal, vertical, etc.). The fourth number in a typical rod designation species the composition of the welding rod covering, or ux, and the suggested power supply for that rod (DCEP, DCEN or AC). These nal two numbers should be read together to inform the welder if the chosen rod is the best t for the application. In the case of an E7018 rod, the 18 indicates a covering composition containing iron powder, lowhydrogen and potassium, as well as indicating that the rod can be used in all positions, and with DCEP or AC power supply. In addition to the 4 standard numbers in a welding rod designation, a sufx can be added to describe a lowalloy steel electrode. In this case, the additional sufx provides information pertaining to the composition of the ller metal. For example, the sufx -A1 indicates the steel rod is a carbon-molybdenum electrode containing 0.5% molybdenum. Table 3.1 lists the available carbon steel covered electrodes dened by the American Welding Society (AWS) with the tensile strengths of 60 and 70 ksi (E60XX and E70XX). This information has been adapted from AWS A5.1-91 (AWS, 1991). The information presented in Table 3.1 includes only those electrodes which have a consumable metal tensile strength of 60 and 70 ksi, as these are the most commonly used electrodes for structural applications. For the repair of historic wrought iron, the use of higher strength electrodes is not necessary, as the strength of the electrode will always exceed the strength of the wrought iron base metal which typically has a tensile strength in the range of 26

PAGE 39

45-55 ksi (Aston & Story, 1939), and is consistent with the testing performed as part of this thesis. Table 3.1: Carbon Steel Covered Arc Welding Electrodes Adapted from AWS A5.1-91 (F: Flat, V: Vertical, OH: Over-head, H: Horizontal) AWS Classication Flux Type Recommended Position Electric Current Type E6010 High cellulose sodium F, V, OH, H DCEP E6011 High cellulose potassium F, V, OH, H AC or DCEP E6012 High titania sodium F, V, OH, H AC or DCEN E6013 High titania potassium F, V, OH, H AC, DCEP, DCEN E6020 High iron oxide H-llets AC or DCEN E6022 High iron oxide F AC, DCEP, DCEN E6027 High iron oxide, iron powder H-llets, F AC or DCEN E7014 Iron powder, titania F, V, OH, H AC, DCEP, DCEN E7015 Low hydrogen sodium F, V, OH, H DCEP E7016 Low hydrogen potassium F, V, OH, H AC or DCEP E7018 Low hydrogen potassium, iron powder F, V, OH, H AC or DCEP E7024 Iron powder, titania H-llets, F AC, DCEP, DCEN E7027 High iron oxide, iron powder H-llets, F AC or DCEN E7028 Low hydrogen potassium, iron powder H-llets, F AC or DCEP E7048 Low hydrogen potassium, iron powder F, OH, H, Vdown AC or DCEP 27

PAGE 40

Gas Metal Arc Welding Another common and fairly simple welding method is the Gas Metal Arc Welding (GMAW) process. In the GMAW process, a continuously fed wire provides the consumable electrode that creates the arc and is melted into the weld. To shield the weld and protect it from impurities, the welding arc is shielded by an inert gas. This process is also commonly known as metal inert gas (MIG) welding, and is used with a constant voltage DC welding machine. Commonly, gases such as argon, carbon dioxide or helium are used as the shielding gas, and are selected for use based on the type of base metal being welded. For the GMAW process, a single insulated cable caries the wire feed as well as the feed for the shielding gas. The GMAW process has potential to be used for wrought iron repair, because for a typical groove weld, a GMAW weld requires a much smaller groove angle for a quality penetration weld. Because of this, less ller metal is required, and less heat is transferred to the base metal. Due to it's simplicity, GMAW is one of the most efcient welding processes, and with multiple wire types and shielding gas types available for use, it provides great potential for the successful repair of historic wrought iron structures. 28

PAGE 41

Gas Tungsten Arc Welding Similar to the GMAW process discussed above, the gas tungsten arc welding (GTAW) process uses a shielding gas to prevent impurities in the weld. However, in the GTAW process a non-consumable tungsten electrode is used, as opposed to the consumable electrodes of the SMAW and GMAW processes. In this process a ller metal may or may not be used, and an AC or DC power source may be used. The GTAW process is commonly referred to as tungsten inert gas (TIG) welding. Inert gases that are commonly used with the GTAW process include, but are not limited to, Argon, Helium, and Argon-Hydrogen mixtures. Compared to the SMAW and GMAW processes, the GTAW process is more difcult to perform correctly, but when properly executed produces extremely clean, high quality welds. For this reason, the GTAW process is used frequently in the welding of pipe applications such as petroleum and chemical pipe lines. Although technically more difcult, the GTAW process has potential to be used to perform high quality repair of historic structural wrought iron. Standard Joint Types In modern welding practice there are ve standard joint types that are commonly used. Although very limited research has been performed using modern welding techniques on historic wrought iron, all ve of the standard joint types have the potential to be used for the repair of historic wrought iron structures, and will be presented below, in hopes that future research will performed using all these joint types with wrought iron. The ve standard joint 29

PAGE 42

types are as follows: butt joint, corner joint, t-joint, lap joint and edge joint. Figure 3.1 is a graphical representation of the ve standard modern welding joint types. Based on the application of each of these joints, and the required strength of the welded joint, several types of welds can be applied to each of these joint types. For some weld types, such as the double-bevel groove weld performed on a butt joint, special preparation of the joint is necessary. For other weld types, such as a llet weld performed on a t-joint, no unique joint preparation is required. Table 3.2 shows an abridged list of various welds that can be performed on each joint type. For this study, the research focused on developing an acceptable welding procedure using a butt joint and a double bevel groove weld (see Figure 4.8). This was the only joint utilized for this research, due to the necessity of changing only one variable at a time, and the iterative nature of this process. See Chapter IV for a description of the welding processes performed in this study. Figure 3.1: Standard Joint Types 30

PAGE 43

Table 3.2: Applicable Welds for Basic Joint Types Adapted from AWS A3.0: 2001 Joint Type Applicable Welds Butt Joint Bevel-groove, Flare-bevel-groove, Flare-V-groove, J-groove, Square-groove, U-groove, V-groove T-Joint Fillet, Bevel-groove, Flare-bevel-groove, J-groove, Squaregroove Edge Joint Bevel-groove, Flare-bevel-groove, Flare-V-groove, J-groove, Square-groove, U-groove, V-groove, Edge Corner Joint Fillet, Bevel-groove, Flare-bevel-groove, Flare-V-groove, Jgroove, Square-groove, U-groove, V-groove Lap Joint Fillet, Bevel-groove, Flare-bevel-groove, J-groove, Squaregroove Weld Quality, Flaws & Testing Repairing historic structural wrought iron with modern welding methods has great potential, but while researching potential welded repair methods, it is imperative that engineers understand the effect that even very small defects can have on welds. The following discussion will be limited to the common defects and problems that are common to welds performed using the SMAW process, as that was the only process used for the research described by this thesis. As described above, electrodes used for shielded metal arc welding are coated with a ux that, when melted during the welding process, oats to the top of the molten weld puddle and hardens as the metal cools. This coating that ends up on top of the welded metal is called slag. For all multi-pass welds, such as those used for the double bevel welds described in Chapter IV of this thesis, the slag must be completely removed with a wire brush and/ or chipping hammer 31

PAGE 44

before the next weld is performed. Any remaining slag that is left can create a slag inclusion, which will leave pockets in the weld, and can signicantly decrease the strength of the welded joint. Slag inclusions are likely to leave the joint itself weaker than the base metal, which in structural practice will create a potential stress concentration that may lead to a sudden failure of the joint. Beyond the easily avoided problem of slag inclusion, there are several factors that will inuence the quality of a weld using the SMAW process. Four signicant factors that are controlled by the welder are, arc length, welding travel speed, width of weld bead, angle and position of electrode. Arc length refers to the distance between the electrode and the base metal, and affects the heat of the weld, the quality of penetration and the purity of the weld. Welding travel speed is the pace at which the welder moves the electrode from one end of the weld to the other. Welding speed that is too fast or slow can signicantly decrease the quality of the weld by causing various defects. The width of the weld bead is varied by the motion of the electrode and the speed of travel. The correct width of the weld must be applied by the welder to ensure that sufcient penetration is provided, but also that excessive heat is not applied to the base metal. The angle of the electrode refers to how the electrode is positioned with respect to the face of the base metal, and can affect the penetration of the weld, and the formation of a proper weld bead. All of these factors must be balanced and adjusted appropriately by the welder based on the joint type selected, the base metal, the electrode being used and the type of weld required. If the correct adjustment is not made, several weld 32

PAGE 45

defects, or aws can occur. Some of these weld defects are easily visible to the naked eye. However, some defects are only discovered through the use of more sophisticated testing methods. Four common defects that can be identied by simple visual inspection are poor welding proportions, undercutting, poor penetration, and surface aws and defects. An experienced welder can not only recognize all of these defects but also recommend appropriate adjustments to the welding process to x them. Flaws and defects that cannot be detected by visual inspection will only be found by destructive and non-destructive testing. Two non-destructive testing methods that are common in the construction industry are ultrasonic and x-ray testing. Both of these methods are used to the verify the quality of welds made in the construction industry and have the potential to be used to investigate welds performed on historic wrought iron. Additionally there are several destructive testing methods used to test the quality of a weld, and that should be used while performing research on welding methods for historic wrought iron. The two most common, and the two used for the research presented by this thesis, are the tensile test and the guided bend test. For a detailed description of these tests see Chapters IV-VI. Welding Procedure Qualication Modern welding procedures used in the structural engineering and construction industry today are governed by the American Welding Society (AWS) Structural Welding Code, AWS D1.1. This standard includes many pre33

PAGE 46

qualied welding procedures intended for use in structural steel construction. Although many of these pre-qualied welding procedures are available for the welding of structural steel, in researching the available literature, the author has found no published, pre-qualied welding procedure for the welding of historic structural wrought iron. The purpose of the research discussed in this thesis is to provide a tested welding procedure to be used on historic wrought iron. My hope is that the welding procedures discussed in this thesis, or similar procedures will eventually be put through an AWS approved welding procedure qualication testing program. The essential variables that must be dened in a pre-qualied welding procedure, and that are applicable to the SMAW process, are the following: base metal thickness, base metal properties, ller metal properties, pre-heat temperature, post-heat temperature, joint geometry, and welding position. In many structural applications, if a type of weld is required that does not t within a pre-qualied welding procedure, a welding procedure qualication record (WPQR) must be submitted and approved prior to the weld being used in a fabrication shop or in the construction eld. For the qualication of a groove weld, specimens must be cut from the welded material and used for tension tests and guided face and root bend tests to verify the strength, and ductility of the nal welded specimen. These tests were performed on the welded wrought iron studied for this thesis, and the testing methods and results can be found in Chapters IV-VI. 34

PAGE 47

CHAPTER IV WROUGHT IRON TESTING TEST DESIGN & CONFIGURATION Base Metal Testing Design & Conguration Test with 220k MTS Before any tests on welded wrought iron were performed, and before any welds were completed, samples of the non-welded wrought iron obtained from Minnesota Bridge #5721 were tested in order to determine basic material physical properties. For this reason, tensile coupon tests were performed on ve base metal specimens. Three of these tests were performed on a 220k MTS testing frame, shown in Figure 4.1. The other two utilized a 20k MTS testing frame, which can be seen in Figure 4.2. Both testing machines are located in the Civil Engineering Laboratory at the University of Colorado Denver. Initially, only the larger of the two testing frames was used, in order to reduce the amount of machining required of each test specimen. The 20k frame was used to verify the results of the larger testing frame and to conrm the accuracy of all obtained results. 35

PAGE 48

Figure 4.1: 220k MTS Testing Frame Figure 4.2: 20k MTS Testing Frame 36

PAGE 49

! For wrought iron samples to be tested using the 220k MTS, a system had to be devised to mount the wrought iron specimen into the testing machine. Unlike the 20k MTS which uses simple clamps to grip the specimen, the 220k MTS utilizes a two inch diameter pin at each head to secure the specimen being tested. The wrought iron material being tested was taken from two different eyebar sections of Minnesota Bridge #5721. The portions of these eye-bars that were cut out and tested were rectangular in cross section, with approximate dimensions of 7/8 in. thick and 2 in. wide. Because these specimens were limited to a width of 2 inches, the 2 inch diameter pin of the 220k MTS could not be directly connected to the wrought iron specimen. To solve this problem, two clamp plates for each head of the MTS were machined from mild steel. Each clamp plate contained a hole cut to t the 2 in. pin, a 5/16 in. deep by 2 in. wide slot to accept the wrought iron at bar specimen, and two 13/16 in. diameter holes to accept the 3/4 in. bolts which were used to clamp the two plates onto the specimen, thus securing the wrought iron specimen through a clamping, friction force. Figure 4.3 below shows the design of the mild steel clamp bars that were fabricated for the purpose of testing the wrought iron samples with the 220k MTS testing frame. To ensure sufcient friction force, and avoid transferring the load of the test into the holes in the wrought iron specimen through bearing, 3/4 in. diameter bolts were used, and a torque of 425 ft-lb was applied to each bolt. 37

PAGE 50

Figure 4.3: Design of Steel Clamp Bar for Tension Test ! The decision to rely on the friction force of the clamp plates was driven by the inability of holes drilled in the wrought iron specimen to withstand the required force through direct bearing action. Limited by the thickness and width of the provided wrought iron samples, had the researchers relied solely on the bearing strength of holes in the wrought iron, the allowable tension force in the specimen would have been signicantly decreased. Using the clamp plate method and relying on the friction force to resist the tension loading from the MTS machine allowed a higher force to be applied and greatly decreased the 38

PAGE 51

required machining of the wrought iron specimen, thus expediting the research process. The wrought iron specimens to be tested were machined to have a smaller cross-sectional area through the middle of the specimen, so the location of failure in the specimen could be limited to a relatively small zone. A crosssectional area of approximately 0.75 in 2 was obtained in the failure zone of the specimen. The cross-sectional area in the failure zone was specied based on the expected ultimate tensile strength of historic structural wrought iron (see Chapter II of this thesis), and a preferred specimen failure in the range of 38,000 to 40,000 pounds of tensile loading. Each wrought iron specimen was cut to be 18 in. long, with a pair of holes drilled at each end to match the holes in the clamp plate assembly. Figure 4.4 shows the design of the wrought iron base metal specimens used in the 220k MTS test frame. For the base metal tension tests, strain gauges were placed in the expected failure zone. These gauges were used for the sole purpose of obtaining an accurate stress-strain diagram, and thereby determining the elastic modulus of the wrought iron material in question. The strain gauge data was collected in tandem with data from the MTS machine until the yielding of the wrought iron specimen occurred, at which point the strain gauges failed and stopped producing usable data. 39

PAGE 52

Figure 4.4: Design of Wrought Iron Tension Test Specimen In addition to the strain gauges, gauge marks were placed on the wrought iron specimen (see Figure 4.4) and the distance between gauge marks was veried with a caliper to the nearest thousandth of an inch. After each specimen was tested to failure, these gauges marks were measured a second time. The difference between the rst and second measurement of the gauge marks provided an accurate measurement of the elongation of the specimen, which 40

PAGE 53

allowed the researchers to calculate the ductility of each tested specimen. Figure 4.5 shows a picture of the complete base metal tension test, with the 220k MTS, steel clamp plates, wrought iron specimen and strain gauges. Figure 4.5: Base Metal Tension Test Setup The tension tests of the base metal specimens were performed using a displacement controlled testing function. The initial tests were performed at a rate of 1/16 in. per minute, but after review of the initial data, it was determined that there would be no negative affects if the rate was increased, and it would greatly expedite the testing process. After the initial test, each tension test was 41

PAGE 54

performed at a rate of 1/8 in. per minute. The computer console used for the control of the 220k MTS was set to provide output data sets of displacement vs. time and force vs. time. Test with 20k MTS The setup for the tests performed on the 20k MTS test frame was very simple because of the conguration of the testing apparatus. As can be seen in Figure 4.2, the 20k MTS has built-in clamps at each head. These clamps are designed with a system of springs which increases the clamping force as the heads move apart. Therefore, as the tension load on the specimen increases, the friction force holding the specimen also increases. The two specimens created for use with the 20k MTS were designed to fail at approximately 15 k, allowing sufcient tolerance to ensure the specimen failed within the range of the testing machine. To ensure failure at this lower tensile load, the specimens were machined to have a rectangular cross-section 3/8 in. thick and 3/4 in. wide. This provided a cross-sectional area of 0.281 in 2 throughout the failure zone. In addition to the displacement and force data recorded from the heads of the testing machine, an extensometer (see Figure 4.6) was placed in the middle of the expected failure zone and was connected to the MTS control computer in order to record strain during the elastic phase of each test. The extensometer was removed prior to yielding of the specimen, in order to prevent damage to the sensitive electronic instruments within the device. 42

PAGE 55

! The tests performed on the 20k MTS used the same set of parameters which were used for the tests performed on the 220k MTS. Each of the two tests was performed at a rate of 1/8 in. per minute. The computer console used for the control of the 20k MTS was set to provide output data sets of time vs. displacement and time vs. force, in addition to the output of strain values during the elastic loading phase. Figure 4.6: Wrought Iron Specimen for 20k MTS with Extensometer The base metal tension tests demonstrated the physical properties of the wrought iron material being studied and were essential to the later testing of welded wrought iron specimens. From the base metal tension tests, the researchers were able to determine the average elastic modulus, yield strength, ultimate strength and ductility of the wrought iron material that is discussed in this thesis. For the results of the base metal tension test, see Chapter V, and for a discussion and analysis, see Chapter VI. 43

PAGE 56

Base Metal Spark Test Design & Conguration In addition to the tension tests described above, the wrought iron specimens taken from eye-bars of Minnesota Bridge #5721 were subjected a spark test. Simple in nature, the spark test is a quick, in-expensive way of identifying the material at hand. Besides the characteristics demonstrated during loading and failure, and visual examination of the failed specimens, the spark test provides a third way to verify that the material being tested is wrought iron and not another metal that has been mis-labeled. The spark test was performed simply by taking a piece of wrought iron that had been cut from one of the Minnesota eye-bars and grinding it on a standard machine shop grinding wheel. In order to document the results of the spark test, the lights were turned off, and photographs were taken while the wrought iron sample was being pressed to the grinder and sparks were being generated. For the results and discussion of the ndings of the spark test see Chapters V and VI of this thesis. Welded Specimen Tension Tests Design & Conguration Prior to the beginning of the welding process, a detailed welding plan had to be developed. The goal of the welding plan was to control as many variables as possible in the welding process, and change only one variable at a time, thereby making it easier after testing to draw conclusions regarding which welding procedures showed better performance. All welds were performed at Western Colorado Community College, Grand Junction, CO, by an advanced 44

PAGE 57

welding student. An eye-bar labeled "M1" was used for all the welded tension tests. This eye-bar was originally referred to as diagonal member U2W-L3W while it was part of Minnesota Bridge #5721. Additionally, three specimens were cut from eye-bar "M5" or diagonal member U2E-L3E. The welded specimens for the tensile tests were prepared by cutting 18.8 in. and 20 in. long sections out of the M1 and M5 eye-bars. Figure 4.7 shows the diagram that was used to direct the cutting of eye-bar M1 in order to maximize the number of usable specimens that could be obtained from the eye-bar. See Figure 4.8 for a diagram of the segmentation of eye-bar M5. Figure 4.7: Eye-Bar M1 Segmentation Diagram Each of these sections (M1-1 through M1-9, and M5-10 through M5-12) were cut in half and each half was prepared with a double bevel on one end. The diagram in Figure 4.8 shows the type of weld that was used to weld together the two halves of each specimen. For all specimens, the dimensions of the landing, 45

PAGE 58

f' and the root opening, R' were held constant at 1/16 in. The groove angles, and # ', were varied equally so that the top groove angle was always equal to the bottom groove angle. In addition to the groove angle, the pre-heat temperature was adjusted for each weld. Both the groove angle and the pre-heat temperature were adjusted based on the welding parameters shown in Table 4.1. For each of these welds, a Lincoln Electric Precision TIG 185 machine (see Figure 4.9) set up for the SMAW, DCEP process was used along with 1/8 in. diameter, E7018 welding rods. The amperage was kept within a range of 115-130 amps, but within that range, it was adjusted at the discretion of the welder (see Table 4.1). Table 4.1: Weld Parameters for Welded Tension Tests Value not provided Bar Number Angle (degrees, = # ) Preheat Temp. ( o F) Amperage (amps) M1-1 30 No Preheat 115 M1-2 30 300 120 M1-3 30 600 120 M1-4 45 No Preheat 130 M1-5 45 300 130 M1-6 45 600 125 M1-7 60 No Preheat 125 M1-8 60 300 125 M1-9 60 600 125 M5-10 75 No Preheat M5-11 75 300 M5-12 75 600 46

PAGE 59

Figure 4.8: Double Bevel Full Penetration Groove Weld Figure 4.9: Lincoln Electric Precision TIG 185 After the completion of the welds on specimens M1-1 through M1-9, each specimen was machined per the sketch shown in Figure 4.4, in order to be used 47

PAGE 60

with the clamp plates in the 220k MTS testing frame. Each specimen was tested using the same parameters used for the base metal tension tests, described at the beginning of Chapter IV. Chapters V and VI contain the results, analysis and discussion of the tension tests performed on the welded specimens. Welded Specimen Bend Tests Design & Conguration The welded specimens used for bend tests were created with a very similar process to those used for the tension tests. For the bend tests specimens, diagonal member U2E-L3E from Minnesota Bridge #5721, re-labeled as M5, was used. This member was cut into several 8 in. long pieces (see Figure 4.10). Each piece, labeled M5-1 through M5-9, was then cut in half and one end of each half was prepared for the double bevel weld shown in Figure 4.8. Figure 4.10: Eye-Bar M5 Segmentation Diagram 48

PAGE 61

! After the completion of the weld, the 8 in. long welded specimen was cut in half longitudinally, and machined to the dimensions required for the bend test apparatus. The required preparation for the nal specimen to be subjected to the bend test is shown in Figure 4.11. This nal cut and machining provided the researches with two specimens (each with a thickness of approximately 3/8 in) for each weld parameter listed in Table 4.2. This allowed a root and a face bend test to be performed for each set of weld parameters. Table 4.2: Weld Parameters for Welded Bend Tests Bar Number Angle (degrees, = # ) Preheat Temp. ( o F) M5-1 30 No Preheat M5-2 30 300 M5-3 30 600 M5-4 45 No Preheat M5-5 45 300 M5-6 45 600 M5-7 60 No Preheat M5-8 60 300 M5-9 60 600 49

PAGE 62

Figure 4.11: Preparation of Bend Test Specimen For the root bend test, the the testing plunger is placed against the face of the weld, so the side of the specimen with the root of the weld is subjected to tensile stress. Similarly, for the face bend test, the specimen is oriented so the face of the weld is subjected to tensile stress. A photograph of a nal prepared specimen, prior to bending can be seen in Figure 4.12. Provisions were made to test the welded specimens using a root bend test and a face bend test, because both tests are required to verify the quality of the weld and for welder qualication. 50

PAGE 63

Figure 4.12: Prepared Bend Test Specimens After all the specimens were prepared and ready for testing, they were transported to the welding shop and testing facilities of Emily Grifth Opportunity School in downtown Denver. The bend tests were performed per testing standard ASTM E290, Standard Test Methods for Bend Testing of Material for Ductility The guided-bend test, per ASTM E290, section 3.6 was used (ASTM E290-97a), and the testing machine was set up to accommodate the 3/8 in. thick specimens. The goal was for each specimen to be bent 180 degrees, or a Type 1 bend as designated by ASTM E290. Figure 4.13 shows the schematic of the guided-bend test that was used for this study. For the bend tests recorded in this thesis, a testing machine equipped with a hydraulic plunger was used. The testing machine used was the Watts W-50 Tensile & Bend Weld Tester, manufactured by 51

PAGE 64

Watts Specialties of Milton, WA. See Figure 4.14 for a photograph of the bend test machine which was used for this study. The results, analysis and discussion of these bend tests can be found in Chapters V and VI of this thesis. Figure 4.13: Guided Bend Test Schematic (Adapted from ASTM E290-97a) 52

PAGE 65

Figure 4.14: Watts W50 Tensile & Bend Weld Tester 53

PAGE 66

CHAPTER V WROUGHT IRON TESTING RESULTS Base Metal Testing Results As stated in the previous chapter, ve base metal samples taken from Minnesota Bridge 5721 were tested. Three base metal samples were tested on the 220k MTS testing frame and two base metal samples were tested using the 20k MTS testing frame. These tests were performed in order to determine the physical properties of the wrought iron material, as well as to provide a basis for comparison with the welded wrought iron specimens. These specimens were taken from eye-bar sections labeled M1 and M3 (see Chapter IV), and were given the the label "B-MX-X," designating them as base metal, not welded, specimens. The following properties of the base metal were determined from each test: yield strength, ultimate tensile strength, ductility (measured by percent elongation) and percent area reduction (measured at location of specimen failure). Additionally from one of the base metal specimen tests (B-M1-2) the elastic modulus was determined from the stress-strain diagram. The elastic modulus was not calculated for every specimen, due to unsuccessful coordination of the strain gauge data with the MTS output data during some of the tests. However, the small variance of the ultimate strength and yield strength between the specimens suggests a consistent elastic modulus for all specimens. Table 5.1 displays the results of the base metal tension tests. Graphical results of test specimens B-M1-1 and B-M1-2 can be seen in Figures 5.1 and 5.2, 54

PAGE 67

respectively. The remaining specimen graphical results can be seen in Appendix A, and relevant photographs can be found in Appendix B. See Chapter VI for a discussion of these test results. Table 5.1 Base Metal Tension Test Results *Indicates value not calculated, or determined to be inaccurate Specimen # F y (ksi) F u (ksi) % Elongation % Area Reduction E (ksi) B-M1-1 34.3 49.6 N/A 35.8 N/A B-M1-2 37.3 48.7 33 40.6 28,100 B-M3-2 37.3 50.7 32 N/A N/A B-M3-3 39.1 56.4 22.5 N/A N/A B-M3-4 33.6 47.4 28.1 42.4 N/A Mean 36.3 50.6 28.9 39.6 Std. Dev. 2.30 3.48 4.76 2.79 CV 0.06 0.07 0.16 0.07 55

PAGE 68

Figure 5.1: B-M1-1 Force vs Displacement Plot 0 4000 8000 12000 16000 20000 24000 28000 32000 36000 40000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 FORCE (LB) DISPLACEMENT (IN) 56

PAGE 69

Figure 5.2: B-M1-2 Stress vs Strain Plot (Data collection stopped after yielding of specimen) 0 10000.0 20000.0 30000.0 40000.0 0 0.0023 0.0045 0.0068 0.0090 Stress (psi) Strain (in/in) 57

PAGE 70

Spark Test Results The spark test, performed to verify the wrought iron nature of the tested material was performed using the shop grinding wheel at the University of Colorado Denver. Because this is a visual test, the results are presented here in photograph form in Figure 5.3. For comparison, a photograph of a spark test performed on mild steel has been included (see Figure 5.4). Reference Chapter VI for a discussion of these results. Figure 5.3: Spark Test Photograph Wrought Iron from MN Bridge 5721 Note: Sparks appear as spurts, without "forks" or "ngers", indicating very low carbon content of metal 58

PAGE 71

Figure 5.4: Spark Test Photograph Mild Steel (Photo by Chomsrimake. Used with permission) Note: Sparks "fork" at ends, indicating higher carbon content Welded Specimen Tension Tests Results The welded wrought iron specimens were tested using only the 220k MTS testing frame with the process described in Chapter IV. The goals for the tension tests on the welded specimens were two-fold; rst to verify that the specimen would fail well away from the heat affected zone (HAZ), and secondly to see if the welding of the wrought iron would signicantly affect the physical properties of the metal. To this end, the same values shown for the base metal specimens were determined for the welded specimens. Additionally, photographs are included (Figures 5.5 and 5.6) to show the location of failure relative to the location of the weld. In both photographs shown below it is clear that the 59

PAGE 72

specimen failed away from the heat affected zone, and as shown in Table 5.2, none of the nine welded specimens failed in the HAZ. The area reduction and elongation, or "necking," typical to all the tested specimens studied in this thesis, can be seen by the photographs in Figures 5.5 and 5.6. This necking occurred in the nal minutes of loading prior to failure, after the yielding of the specimen. It is of interest to note that many of the welded specimens experienced this necking phenomenon on either side of the weld, and not just in a single location (see Figure 5.6). See Appendix A for "displacement vs force" of all the welded specimen tension tests. Additional photographs can be found in Appendix B. Tables 5.3 and 5.4 show the average yield strength, ultimate strength and percent elongation organized based on the parameters of the weld that were adjusted during the research process: groove angle and preheat temperature. The discussion pertaining to the results of the welded specimen tension tests can be found in Chapter VI. Reference Table 4.1 for associated groove angle and preheat. 60

PAGE 73

Table 5.2 Welded Specimen Tension Test Results Specimen # F y (ksi) F u (ksi) % Elongation % Area Reduction Failure in HAZ M1-1 32.8 48.5 26 38.5 No M1-2 32.7 50 23.8 25.8 No M1-3 28.4 45.1 26.5 24 No M1-4 31.1 50.1 24.5 24.9 No M1-5 33.1 51.5 25.7 26.5 No M1-6 33.0 50.7 25.6 26.7 No M1-7 34.7 52.7 25.7 27.2 No M1-8 31.8 48.6 24.9 24.5 No M1-9 32.6 50.7 26.3 25.8 No M5-10 30.8 48.1 24.0 36.8 No M5-11 31.4 49.4 23.9 34.7 No M5-12 32.6 48.2 21.1 37.4 No Mean 32.1 49.5 24.8 29.4 Std. Dev. 1.56 1.97 1.51 5.64 CV 0.05 0.04 0.06 0.19 Table 5.3 Welded Specimen Tension Test Results By Groove Angle (Values are averages of all welds performed with noted groove angle) Groove Angle F y (ksi) F u (ksi) % Elongation 30 31.3 47.9 25.4 45 32.4 50.8 25.3 60 33.0 50.7 25.6 Std. Dev. 0.877 1.646 0.184 CV 0.027 0.033 0.007 61

PAGE 74

Table 5.4 Welded Specimen Tension Test Results By Preheat Temperature (Values are averages of all welds performed with noted preheat temperature) Preheat Temp. ( o F) F y (ksi) F u (ksi) % Elongation None Room Temp. 32.9 50.4 25.4 300 32.5 50.0 24.8 600 31.3 48.8 26.1 Std. Dev. 0.806 0.833 0.668 CV 0.025 0.017 0.026 Figure 5.5: Welded Specimen Tension Test Photograph 1 62

PAGE 75

Figure 5.6: Welded Specimen Tension Test Photograph 2 Welded Specimen Bend Tests Results In addition to the tension tests, guided bend tests were performed on nine welded specimens. Each weld type was tested by a face bend and a root bend test (see Chapter IV). The goal was for each specimen to be bent 180 degrees without signicant visual faults such as cracking or splitting. Of the nine root bend tests that were performed, eight were successful, while one specimen broke well before reaching the desired 180 degree bend. The face bend tests were not as successful: all nine specimens subjected to the face bend test failed prior to obtaining the required bend. The results organized by specimen number can be seen in Table 5.5. Selected photographs of a successful and an un-successful 63

PAGE 76

specimen can be seen in Figures 5.7 and 5.8. Additional photographs can be seen in Appendix B. The results of the welded specimen bend tests are discussed in Chapter VI. Reference Table 4.2 for associated groove angle and preheat. Table 5.5 Welded Specimen Bend Test Results Specimen # Root Bend Test Result Face Bend Test Result M5-1 Pass Fail M5-2 Pass Fail M5-3 Pass Fail M5-4 Fail Fail M5-5 Pass Fail M5-6 Pass Fail M5-7 Pass Fail M5-8 Pass Fail M5-9 Pass Fail 64

PAGE 77

Figure 5.7: Welded Specimen Successful Root Bend Test (Specimen M5-1) Figure 5.8: Welded Specimen Failed Face Bend Test (Specimen M5-3) 65

PAGE 78

CHAPTER VI DISCUSSION FROM TEST RESULTS Base Metal Testing Discussion The tension tests of the base metal specimens taken from Minnesota Bridge 5721 demonstrated good strength and ductility. The average yield strength of 36.3 ksi and average ultimate tensile strength of 50.6 ksi (reference Table 5.1) are within the range of expected values based on comparison with all reviewed published data on the physical properties of wrought iron (reference Chapter II). Verifying the strength of this historic wrought iron is crucial to the ability of engineers to condently recommend repairs for the metal. If repair methods that develop the full strength of the original material are used, engineers can be condent that the age of the material will typically not have an affect on the tensile strength of the material. Although the values of yield strength and ultimate strength are within the expected range, it is signicant to note that this wrought iron manufactured approximately 130 years ago, and in constant use for most of that time, maintained it's physical properties so well, and was not affected by corrosion or fatigue. The low coefcient of variation (CV) demonstrated in the yield strength and ultimate strength of these base metal samples displays the fairly constant nature of these properties. This is consistent with what has been seen by other researchers, such as Gordon and Knopf (reference Chapter II) who note that slight chemical variations in wrought iron will affect ductility of the material much more than they will affect the strength of the 66

PAGE 79

material (Gordon & Knopf, 2005). In other words, it is to be expected that the values of F y and F u for multiple wrought iron samples will have little variance. The variability of wrought iron ductility can be seen in the results of the base metal tension tests. The CV calculated for the measurement of percent elongation is more than two times the CV calculated for the values of F y and F u Additionally, the ductility of the wrought iron tested for this study, is very different from the ductility reported by Bowman (reference Chapter II). In the Purdue report, Bowman described the noticeable lack of ductility in their specimens demonstrated by the lack of any visible deformation or necking of the specimen prior to tensile failure. As can be seen in various photographs (reference Figure 6.1 and Appendix B) each specimen tested for this study went through a visible necking stage (plastic deformation) prior to failure. Beyond the visible differences in ductility between the specimens tested here and those tested by Bowman in the Purdue study, the recorded percent elongation of each study is signicantly different. The average elongation of 11.7% recorded by Bowman is nearly 2.5 times less than the average elongation calculated by the testing performed in this study (28.9%, reference Table 5.1). Although no chemical analysis was performed on the wrought iron tested by this study, it can be deduced from the high ductility of the specimens, that this wrought iron has a relatively low phosphorus content. As stated in Chapter II of this thesis, an increase in the phosphorus content of wrought iron will decrease ductility. This is further demonstrated by the lack of ductility in the specimens of the Purdue study, which did undergo chemical analysis and were shown to have a phosphorus content 67

PAGE 80

between "0.25 and 0.36 percent by weight" (Bowman, 2004). This phosphorus content is nearly three times larger than that shown in the typical wrought iron chemical analysis of Table 2.1. The results of the wrought iron base metal specimen tension tests demonstrate consistent yield strength and ultimate strength, as well as demonstrating high ductility. When compared to published data provided by previous wrought iron testing, the values determined from the base metal tension tests performed in this study, support the conclusion that chemical variations in wrought iron can have signicant impact on the ductility of the material, while not greatly affecting the yield strength or ultimate tensile strength. Figure 6.1: Necking Base Metal Tension Test 68

PAGE 81

Spark Test Discussion. The spark test was performed as an inexpensive, efcient method of verifying the low carbon content, and wrought iron nature of the historic metal which was being tested. Reference Figure 5.3 for a photograph of the spark test. The diagram in Figure 6.2 is a sketched depiction of the expected spark patterns for various types of ferrous metals. It can be seen from items B D in the diagram that as carbon content in ferrous metal increases, the amount of "forking" or splitting at the end of the sparks will increase. Therefore, the low carbon content of wrought iron should produce fairly straight line sparks, and no forking at the end of the sparks, as shown in item A of Figure 6.2. Figure 6.2: Spark Test Diagram (Oberg & Jones, 1918) A: Wrought Iron, B: Mild Steel, C: Steel w/ 0.5% 0.85% carbon, D: High-carbon steel, E: High-speed steel, F: Manganese steel, G: Mushet Steel, H: Special magnet steel 69

PAGE 82

! The spark test performed on the wrought iron samples from Minnesota Bridge 5721 clearly demonstrate the spark pattern expected for historic wrought iron. Based on this test, and the observed silica slag, it can be condently stated that the material used for this study is wrought iron. Welded Specimen Tension Tests Discussion As described in Chapter V, welded specimens subjected to tension tests displayed deformation and necking very similar to the un-welded base metal specimens. Additionally, many of the welded specimens displayed this necking on both sides of the weld, prior to failure occurring on one side of the weld. The average yield strength of the welded specimens was 32.2 ksi and the average ultimate strength was 49.8 ksi. Comparing these values to the averages obtained from the base metal samples (Fy = 36.3 ksi, Fu = 50.6 ksi), a slight decrease can be noted (11% for the yield strength and 1.6% for ultimate strength). However, this slight decrease does not mean that the welding of the wrought iron decreased the yield or ultimate strength. The decrease in strength was most likely a result of slight differences in the properties of the wrought iron, as material from two different eye-bars was used for this process. The consistent failure location outside the heat affected zone demonstrates that the strength of the nal welded specimen was not negatively affected by the welding process. The ductility of the welded specimens must also be compared with the ductility shown by the base metal specimens. The results in Tables 5.1 and 5.2 show that the average percent elongation calculated from the base metal 70

PAGE 83

specimens and welded specimens were 28.9% and 25.4%, respectively. Based on these values, a 12% decrease in the average values of percent elongation was seen. This decrease in ductility between the welded and non-welded specimens should was expected because the weld metal did not yield, and that portion of the sample was not subject to any deformation. The fact that the average percent elongation of the welded wrought iron specimens was over 25% leads to the conclusion that the double bevel groove weld which was performed did not have a signicant impact on the ductility of the metal. In addition to the overall results of all welded specimens, Tables 5.3 and 5.4 display the average values for F y F u and percent elongation, organized by weld parameter. Table 5.3 displays the average results of each groove angle used, and Table 5.4 the average results of each preheat temperature used. The values displayed in these tables make it clear that there is no measurable difference in the strength or ductility of the sample welded with 30, 45 or 60 degree bevel angles, or of those in which preheat temperatures were used. It appears that the specimens welded with a 600 o F preheat have a slightly lower F y and F u while showing slightly greater ductility. However, the small differences in these values compared to those of the specimens welded with no preheat or 300 o F preheat prevent hard conclusions to be drawn regarding the benet or detriment of varying preheat temperatures. Both Tables 5.3 and 5.4 show the coefcient of variation to be extremely small, which leads to the conclusion that all nine weld types performed (see Table 4.2) are satisfactory for developing the full strength and ductility of the historic wrought iron. 71

PAGE 84

! Although no signicant difference between the different types of welds can be seen from the testing results, the welding process must be taken into account when recommending a new welding procedure. Because there are so many factors that can adversely affect the weld, it is important that a welding procedure provide as much opportunity as possible for consistent quality. If the required welding procedure is easier to perform, the likelihood of weld defects and aws will be greatly decreased. For the welds performed in this study, it was determined that a 30 degree groove angle is extremely narrow for a standard electrode, and obtaining full penetration on the root weld pass is difcult. According to the individuals performing the welds, even with a 45 degree groove angle full penetration and manipulation of the welding rod was difcult, and the 60 degree groove angle provided the best opportunity for a successful weld. Regarding preheat temperature it was determined by the welders that using a preheat temperature of 300 o F provided the best welding environment, as applying a 600 o F preheat seemed to create too much preheat in the metal and caused the welds to be more difcult to perform. See Chapter VII for suggestions for future research and conclusions based on the research performed here. Welded Specimen Bend Tests Discussion The results of the guided bend tests of the welded specimens present an intriguing challenge to the goal of developing an acceptable welding procedure for historic structural wrought iron. Of the nine specimens tested with the root bend test, one specimen failed, and eight passed. However all nine of the 72

PAGE 85

specimens tested with the face bend test failed prior to reaching the desired 180 degree bend. Reference Chapter V for a summary of these results. In standard welding practice and quality control, the failure of a guided bend test indicates a aw or defect in the weld. However, it is highly unlikely that poor welding was the cause of failure for all of the face bend test specimens. The failure of the face bend tests and the success of the root bend tests is the same phenomenon that was observed by Chomsrimake in his research (reference Chapter II). The hope of this study was that performing welds at various preheat temperatures and with various groove angles would lead to more successful results of the face bend tests. Each of the nine failed specimens had very similar failure locations. All of the face bend test failures occurred at the edge of the fusion zone, where the weld metal and the base metal join. See Figure 5.7 for a photograph of this type of failure. From the results obtained here, it is clear that the bend test failures were independent of the preheat temperature and the groove angle. While the failure of the bend tests does mark the need for continued research into these wrought iron welding procedures, it is important to note that the bend tests are required only for welder qualication. The tension tests performed on the welded wrought iron specimens demonstrate that the welding procedure described in Chapter IV fully maintains the physical characteristics of the historic metal. 73

PAGE 86

CHAPTER VII CONCLUSIONS & RECOMMENDATIONS Conclusions from Research The wrought iron specimens taken from Minnesota Bridge 5721 provided important information and aided to the effort of providing acceptable repair methods for historic structural wrought iron. The series of tests performed on the base metal and the welded wrought iron specimens yielded valuable data and provide valuable insight to the ongoing research of the repair of wrought iron structures. The spark tests performed on the wrought iron samples gave the researchers condence that the metal being tested was indeed historic wrought iron, even without the opportunity to perform detailed chemical analysis. The base metal tension tests provided crucial data dening the physical properties of the historic metal, including yield strength, ultimate strength, ductility and modulus of elasticity. These recorded physical properties were compared with and veried against values found in previously published data on historic wrought iron. The base metal tension tests clearly demonstrated the remarkable and longlasting strength and ductility of historic structural wrought iron. The longevity of this structural material is a key reason why many historic wrought iron structures are still in use today, and why tested and published repair methods are such a necessity for historic preservation. While the base metal tests provided necessary data, and valuable information, the goal of the research described in this thesis was the 74

PAGE 87

development of an acceptable welding procedure for historic structural wrought iron using modern welding techniques. The welding plan used for this research was created with efciency and weldability in mind. The use of the double bevel weld reduced the amount of heat required compared to a single bevel weld, and the parameters of groove angle and pre-heat were varied with the goal of determining the most effective welding procedure for the historic wrought iron metal. The tension tests performed on the welded wrought iron specimens, and the recorded results, demonstrate that variations in the welding procedures did not signicantly impact any of the measured physical properties of the wrought iron. As discussed in Chapter VI of this thesis, the data from the tension tests did not reveal any signicant difference in the structural properties of the specimens welded with various preheat temperatures and groove angles. However, with weldability and quality in mind, it is recommended that a groove angle of 60 degrees be used with a preheat temperature of 300 o F, based on this study. These parameters will give the welder the best opportunity to perform a high quality weld. The success of the tension tests of the welded wrought iron specimens clearly shows that the welding procedures specied in this thesis (reference Chapter IV) will develop the full strength of the base metal. Although all but one of the welded specimens passed the root bend tests, unfortunately, all nine of the welded specimens failed the face bend test (reference Chapter V). The consistent failure of the face bend tests prevents these specic welding procedures from being recommended as pre-qualied welding procedures (reference Chapter III). However, the data collected from all 75

PAGE 88

of these tests does provides valuable information for the ongoing effort to develop a pre-qualied welding procedure for historic structural wrought iron. The following section of this thesis provides suggestions for future research based on the ndings of the research performed. Recommendations for Future Research Welding Methods, Materials There are many welding methods that have the potential to be used for the repair of historic wrought iron structures. As discussed in Chapter III, both GMAW and GTAW methods have characteristics that could enable them to be used efciently for the welding of wrought iron. However, there is still research to be done regarding the use of SMAW welding for wrought iron repair. All the previously conducted research discussed in this thesis, as well as the tests performed for this study, utilized E7018 electrodes. It is my hope that in the future, research will be conducted using multiple electrode types. Better bend test results might be obtained using electrodes with different chemical composition and material properties. Besides research with multiple types of electrode, research should be performed to investigate the effect of post weld heat treatment (PWHT) on welded wrought iron. PWHT is used in some modern welding applications in order to reduce the potential for hydrogen induced cracking, as well as to relieve residual stresses that are present in the metal. Post weld heat treatment of 76

PAGE 89

welded wrought iron may relieve some stresses present in the heat affected zone and allow the bend tests to be successful. Joint Types Although only the butt joint with a double bevel weld was used for this research, it is my hope that further studies will be performed using many other joint types, in order to provide engineers and contractors with many available options for the welded repair of historic wrought iron structures. The joint types that are most commonly used for modern welding in the construction industry, such as lap joints, and T-joints (reference Figure 3.1), should especially be investigated for their potential to be used in pre-qualied welding procedures for historic wrought iron. Currently, wrought iron welding research has only been performed using V-groove bevel welds, but hopefully, future research will include wrought iron welding procedures involving llet welds on various joint types. Wrought Iron Material Investigation In addition to research investigating welding procedures for historic structural wrought iron, there is a continued need for research into the chemical and physical properties of various historic wrought iron structures. Detailed chemical and physical analysis should be performed on samples of wrought iron from bridges that are still in use. This would greatly increase the condence with which engineers could design retrot and renovation plans for these bridges. The varying nature of wrought iron used in historic structures throughout the world 77

PAGE 90

(reference Chapter II) creates the need for detailed study of the wrought iron material in a structure prior to repairs being made. It is my hope that full chemical analysis, such as that shown in Table 2.2, will eventually be performed on the wrought iron material taken from Minnesota Bridge 5721. The research on the ability of double bevel V-groove welds to repair historic wrought iron taken from Minnesota Bridge 5721 proved to be a valuable step towards the goal of developing a pre-qualied welding procedure for use in the repair of historic wrought iron structures. 78

PAGE 91

REFERENCES Aston, J. & Story, E. (1939). Wrought Iron: It's Manufacture, Characteristics and Applications (2nd Edition), A.M. Byers & CO, Pittsburgh, PA AWS (1991). Specication for Carbon Steel Electrodes for Shielded Metal Arc Welding: An American National Standard, American Welding Society, Miami, FL AWS (2001). Standard Welding Terms and Denitions, American Welding Society, Miami, FL Bowman, M. & Piskorowski, A. (2004). "Evaluation and Repair of Wrought Iron and Steel Structures in Indiana." Joint Transportation Research Program, West Lafayette, IN Chomsrimake, P. (2012) "Arc Welding Procedure for Repairing Wrought Iron in Historic Bridges." M.S. thesis, University of Colorado Denver, Denver, CO Gordon, R. & Knopf, R. (2005). "Evaluation of Wrought Iron for Continued Service in Historic Bridges," Journal of Materials in Civil Engineering, 17(4), 393-399 Johnson, R. & Olson, S. (2012). "Collaboration and Innovation Lead to 3rd Service Life for an Iconic Iron Bridge." Structure Magazine 19(10), 29-31 Mills, A. & Hayward, H. (1922). "The Manufacture of Wrought Iron." Materials of Construction: Their Manufacture and Properties (2nd Edition), John Wiley & Sons, New York, NY, 25-29 Oberg, E. & Jones, F. (1918), Iron and Steel (1st Ed) The Industrial Press, New York, NY 79

PAGE 92

Appendix A Tension Test Result Graphs 0 4000 8000 12000 16000 20000 24000 28000 32000 36000 40000 0 0.5 1.0 1.5 2.0 Force (lbs) Displacement (in) Figure A.1: Specimen B-M3-2 Force vs. Displacement Plot 80

PAGE 93

0 4000 8000 12000 16000 20000 24000 28000 32000 36000 40000 0 0.5 1.0 Force (lb) Displacement (in) Figure A.2: Specimen B-M1-1 Force vs. Displacement Plot 81

PAGE 94

0 4000 8000 12000 16000 20000 24000 28000 32000 36000 40000 0 0.50 1.00 1.50 Force (lb) Displacement (in) Figure A.3: Specimen B-M1-2 Force vs. Displacement Plot 82

PAGE 95

0 1500 3000 4500 6000 7500 9000 10500 12000 13500 15000 0 0.50 1.00 1.50 Force (lb) Displacement (in) Figure A.4: Specimen B-M3-3 Force vs. Displacement Plot 83

PAGE 96

0 1500 3000 4500 6000 7500 9000 10500 12000 13500 15000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.5: Specimen B-M3-4 Force vs. Displacement Plot 84

PAGE 97

0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1 1.5 Force (lbs) Displacement (in) Figure A.6: Specimen M1-1 Force vs. Displacement Plot 85

PAGE 98

0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1 1.5 Force (lbs) Displacement (in) Figure A.7: Specimen M1-2 Force vs. Displacement Plot 86

PAGE 99

0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1 1.5 Force (bs) Displacement (in) Figure A.8: Specimen M1-3 Force vs. Displacement Plot 87

PAGE 100

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 0 0.5 1 1.5 Force (lbs) Displacement (in) Figure A.9: Specimen M1-4 Force vs. Displacement Plot 88

PAGE 101

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.10: Specimen M1-5 Force vs. Displacement Plot 89

PAGE 102

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.11: Specimen M1-6 Force vs. Displacement Plot 90

PAGE 103

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.12: Specimen M1-7 Force vs. Displacement Plot 91

PAGE 104

0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.13: Specimen M1-8 Force vs. Displacement Plot 92

PAGE 105

0 6250 12500 18750 25000 31250 37500 43750 50000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.14: Specimen M1-9 Force vs. Displacement Plot 93

PAGE 106

0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1.0 1.5 Force (lbs) Displacement (in) Figure A.15: Specimen M5-10 Force vs. Displacement Plot 94

PAGE 107

Figure A.16: Specimen M5-11 Force vs. Displacement Plot 0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1 1.5 Force (lbs) Displacement (in) 95

PAGE 108

Figure A.17: Specimen M5-12 Force vs. Displacement Plot 0 5000 10000 15000 20000 25000 30000 35000 40000 0 0.5 1 1.5 Force (lbs) Displacement (in) 96

PAGE 109

Appendix B Photographs Tension Tests Figure B.1: Tension Test with 220k MTS Figure B.2: Tension Test Specimen 97

PAGE 110

Figure B.3: Tension Test with 220k MTS ! 98

PAGE 111

Figure B.4: Tension Test with 20k MTS 99

PAGE 112

Appendix C Photographs Bend Tests Figure C.1: Prepared Bend Test Specimens Figure C.2: Bend Test Operation 100

PAGE 113

Figure C.3: Failed Face Bend Test Figure C.4: Successful Root Bend Test 101