Get documents about "Effect of GMAW on the Mechanical Properties of In-Line Galvanised
W
Description
Effect of GMAW on the Mechanical Properties of In-Line Galvanised
Document Sample
EFFECT OF GMAW ON THE MECHANICAL PROPERTIES OF IN-LINE GALVANISED
COLD-FORMED STEEL
Tim Wilkinson and Gregory J Hancock
Abstract
This paper describes tensile tests to examine the effect of various welding procedures on the
mechanical properties of in-line galvanised cold-formed steel. Cold-formed flats in either Grade C350
or Grade C400 were butt welded. The dip transfer method or the spray transfer method were used
to provide a variety of heat inputs, and two different welding electrodes, Autocraft LW1 and Autocraft
Mn-Mo, were used. The thicker (8 mm) Grade C350 steel displayed a small reduction in the yield and
ultimate stresses when welded compared to the unwelded steel. The 3.8 mm Grade C400 samples
displayed a more significant drop in yield and ultimate stresses when welded, and the drop in strength
was greater when the higher heat input spray method was used. On some occasions the ultimate
strength of the welded 3.8 mm specimens dropped below the yield stress of the parent material.
Keywords
Cold-formed steel, GMAW, butt welds, fillet welds, heat affected zone, mechanical properties, yield
stress, heat input, structural steel, galvanised steel.
Author Details
Dr Tim Wilkinson is a Lecturer, and Professor Greg Hancock is the BHP Steel Professor of Steel
Structures, Centre for Advanced Structural Engineering, Department of Civil Engineering, The
University of Sydney, Sydney, NSW, 2006, Australia.
IIW (International Institute of Welding) Asian Pacific Congress, Paper No. 35, (Published by Welding Technology
Institute of Australia - WTIA), Melbourne, Australia, November 2000.
1
1 INTRODUCTION
A recent innovation in steel products in Australia is the DuraGal range of cold-formed in-line
galvanised hollow and open sections produced by BHP Structural and Pipeline Products (formerly
known as Tubemakers) [1]. The typical steel strip used in the manufacturing process has a nominal
yield stress (fy) of 300 MPa. After cold-forming, the final product has a nominal yield stress in the
range 350 - 400 MPa, depending on the exact process and the shape of the product.
During the welding process, the grains of the cold-worked steel recrystallise, and the heat affected
zone will soften compared to the cold-formed hardness [2]. Consequently, the ultimate tensile
strength (fu) in the heated affected zone (HAZ) may be less than the yield stress of the parent
material.
There are several instances in which a steel structure has to demonstrate ductile behaviour. In plastic
design, the plastic hinges must rotate sufficiently for moment redistribution to take place in the
structure, in order to obtain the strength increase afforded by plastic design. For seismic design,
deformation capacity is essential to dissipate the energy caused by the earthquake motion. In such
cases, the joints of a steel structure are required to show ductile behaviour.
However, if there is a small HAZ in a welded joint, where the ultimate tensile strength is less than the
yield stress in the adjacent unaffected steel, the HAZ will fracture before significant plastic
deformations occur near the joint. This renders the structure unsuited for plastic design or seismic
applications. A previous investigation [3, 4] examining the suitability of portal frame knee joints for
use in a plastically designed structure constructed from cold-formed rectangular hollow sections
(RHS), found that under opening bending moment, the connection fractured in the HAZ before large
plastic deformation occurred. It should be noted that the connection displayed adequate strength, as
opposed to ductility, which means that it was still suitable for use in elastic design.
This paper describes the initial portion of a research project examining the strength and ductility of
joints constructed from in-line galvanised cold-formed steel. The aim is to quantify the effect of
welding on the mechanical properties of cold-formed steel.
2 TEST SPECIMENS
A typical portal frame and possible knee joint details are shown in Figure 1. Two possible connections
are shown. The first is an unstiffened connection in which adjoining RHS are butt welded directly
together. The second detail is a stiffened connection, in which a plate is inserted between the two
RHS legs. Each RHS is either butt welded or fillet welded to the stiffening plate. These two
connection details are suggested by CIDECT [5]. This type of mitred welded connection could also
be used for open steel sections.
Rather than testing a large and expensive connection, it is possible to test a small component of such
a system. Hence, two flat plates were butt welded together to simulate a portion of a more
complicated connection, as shown in Figure 2. The plates would then be tested in tension.
2
Knee Joint
Stiffening plate welded Two legs welded directly
between two legs together
(a) Stiffened welded connection (b) Unstiffened welded connection
Figure 1: Types of connections in a portal frame
Steel strip
Butt weld
Figure 2: Butt weld between 2 DuraGal flats
2.1 Steel Properties
DuraGal flats in nominal thicknesses of 3.8 mm and 8.0 mm were chosen. The 3.8 mm sections were
Grade C400 (fyn = 400 MPa, fun = 450 MPa), and the 8.0 mm flats were Grade C350 (fyn = 350 MPa,
fun = 400 MPa). Currently, there is no Australian Standard applicable to the manufacture of cold-
formed open profiles, and hence the sections are manufactured to an internal BHPSPP Specification,
TS100. The carbon equivalent of this steel is typically CE = 0.24 6 0.30 [1].
2.2 Weld Metal Properties
Two types of welding wire were used in the GMAW process. Autocraft LW1 (fyn = 390 MPa,
fun = 500 MPa) and Autocraft Mn-Mo (fyn = 530 MPa, fun = 630 MPa), to AS/NZS 2717.1 [6] were used.
More details on the wire properties can be found in [7].
2.3 Typical Welding Parameters
Two methods of GMAW were employed, the “dip-transfer” and “spray-transfer” modes. Generally,
the spray transfer method requires a higher wire speed and higher current, and consequently a higher
heat input. It is not possible to include the full details of all welding procedures in this paper, due to
length requirements, however some typical welding parameters are given in Table 1.
3
Welding Machine Transmig 250, with Transmatic 62 wirefeeder
Gas Argoshield 50; 23 % CO2, 77 % Ar; 25 L/min
Electrode Autocraft LW1 or Autocraft MnMo
Dip Method Potential: 18 - 19 V
(range of values) Current: 120 - 135 A
Wire speed: 3700 mm/min
Welding speed: 260 - 400 mm/min
Heat input: 0.334 - 0.494 kJ/mm
Spray Method Potential: 26 V
(range of values) Current: 205 - 210 A
Wire speed: 9700 mm/min
Welding speed: 450 - 920 mm/min
Heat input: 0.412 - 0.728 kJ/mm
Table 1: Typical welding parameters
2.4 Joint Preparation
Square butt welds, and single-V butt welds, welded both sides were used, as shown in Figures 3
and 4.
1
Plate
2
thickness, t
Gap, G Either 1 or 2 welding runs
Typical dimensions:
t = 3.8 mm, G = 0.0 or 0.9 mm
Figure 3: Typical Joint Preparation for 3.8 mm DuraGal Flat
(Type B-C 1a or Type B-C 1b from Table 4.4(A) of AS/NZS 1554.1: 1995 [8])
θ 2 3
1
Plate 4
thickness, t
Either 4 or 5 welding runs
Gap, G Root face, Fr
Typical dimensions:
t = 8.0 mm, Fr = 1.0 mm, θ = 90E
G = 0.0 or 0.9 mm
Figure 4: Typical Joint Preparation for 8.0 mm DuraGal Flat
(Type B-C 2a from Table 4.4(A) of AS/NZS 1554.1: 1995 [8])
4
3 TEST PROCEDURE
Two 150 mm long plates were butt welded together. The sections were either 100 mm wide (3.8 mm
thick specimens) or 150 mm wide (8 mm thick sections). Different tests were performed in
accordance with AS 2205.1 [9].
A tensile coupon was cut longitudinally from the plate in accordance with AS 2205.2.1 [10] as shown
in Figure 5. The butt weld was located transversely at the middle of the coupon. The tensile coupons
were prepared and tested to AS 1391 [11]. An extensometer was used to measure strain. The
coupons were tested in a 300 kN capacity SINTECH Testing Machine with friction grips to apply the
loading as shown in Figure 6. A constant strain rate of approximately 1.0 × 10-3 s-1 was used. In some
cases the weld reinforcement was removed so that a completely flat coupon was tested. In the
remaining cases the weld reinforcement remained.
The properties of the weld metal itself were obtained by performing an all-weld-metal tensile test to
AS 2205.2.2 [12]. Properties of the unwelded steel were determined to AS 1391.
Macro specimens were also cut from the specimens. However the results of the macro section
examination and Vickers Hardness tests are not presented in this paper.
25
Separate
flatbar
45 sections
12.5
Butt
weld
180 90 65
12.5
Extensometer
12.5 (connected to
computer data Test specimen
logger)
45 Butt weld
Figure 5: Dimensions and location of tensile
coupon within welded plate
(all dimensions in millimetres) MTS Sintech grip
Figure 6: Schematic details of tensile testing
5
4 RESULTS
From each test it is possible to obtain stress strain curves such as the one shown in Figure 7.
Figure 7 illustrates how the tensile behaviour changes from the unwelded material, to the welded
sample and weld metal only. The stress is calculated as the load divided by the flat area of the
coupon, and similarly the strain is the ratio of the extension of the specimen measured by the
extensometer divided by the original 50 mm gauge length of the extensometer. However, in the cases
of the specimens where the reinforcement was not removed, the stress and strain are not uniform
along the entire 50 mm gauge length, since the extensometer straddles the reinforced region.
The values of yield stress and ultimate stress are given in Table 2, and can also be seen in Figures 8
to 11. Since the steel is cold-formed there is no well-defined yield stress, and the yield stress quoted
is the dynamic 0.2% proof stress. The term dynamic is used since the stress was determined while
the testing machine was loading at a constant rate of stroke.
Figures 12 and 13 show a selection of fractured specimens.
800
700
600
Stress (MPa)
500
400
Mn Mo Weld Metal
300
3.8 mm Parent Metal
200 MnMo 3.8 mm spray no gap
100 MnMo 3.8 mm dip 0.9 gap
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Strain
Figure 7: Typical stress strain curves
700 600
600 500
500
Stress (M
Stress (MPa)
400
400
300
300
200 Ultimate Stress (fu)
200 8 mm DuraGal Flatbar
3.8 mm DuraGal Flatbar Ultimate Stress (fu) Butt Welded LW1
100 Yield Stress (fy)
100 Butt Welded LW1 Yield Stress (fy) reinforced
Spray no
0
Parent
Dip gap
0
Dip no
Nominal
Dip gap
Weld
Nominal
Dip no
Parent
Spray gap
Dip gap
reinforced
reinforced
Spray no
Weld
Spray gap
gap
Dip gap
gap
gap
gap
Figure 8: Results - 3.8 mm steel, LW1 Figure 9: Results - 8 mm steel, LW1
800 800
700 3.8 mm DuraGal Flatbar 700 8 mm DuraGal Flatbar
Butt Welded MnMo Butt Welded MnMo
600 600
Stress (MPa)
Stress (MPa)
500 500
400 400
300 300
Ultimate Stress (fu) 200 Ultimate Stress (fu)
200
Yield Stress (fy) 100 Yield Stress (fy)
100
0
0
Nominal
Parent
Spray no
Spray no
Dip gap
reinforced
Weld
MnMo
gap reo
MnMo
MnMo
Dip gap
MnMo
Nominal
Parent
Spray no
Spray no
Weld
Dip gap
Dip gap
MnMo
MnMo
gap reo
gap
MnMo
MnMo
reo
gap
Figure 10: Results - 3.8 mm steel, Mn-Mo Figure 11: Results - 8 mm steel, Mn-Mo
6
Material Weld Type Reinforcement Electrode Type Heat Inputa Yield Ultimate Failure locationd
(kJ/mm) Stressb Stressc
(MPa) (MPa)
All weld metal Nominal propertiese - LW1 390 500 -
All weld metal Measured properties - LW1 419 521 -
All weld metal Nominal propertiese - MnMo 530 630 -
All weld metal Measured properties - MnMo 526 699 -
3.8 mm DuraGal Nominal properties - - Unwelded 400 450 -
3.8 mm DuraGal Measured properties - - Unwelded 525 601 -
3.8 mm DuraGal Dip - Gap No LW1 0.365/0.365 392 497 Weld
3.8 mm DuraGal Dip - Gap Yes LW1 0.365/0.365 457 549 HAZ
3.8 mm DuraGal Dip - No Gap No LW1 0.365/0.365 395 490 Weld
3.8 mm DuraGal Spray - Gap No LW1 0.533 369 502 HAZ
3.8 mm DuraGal Spray - Gap Yes LW1 0.533 422 539 HAZ
3.8 mm DuraGal Spray - No Gap No LW1 0.533/0.348 344 487 HAZ
3.8 mm DuraGal Dip - Gap No MnMo 0.351/0.351 435 546 HAZ
3.8 mm DuraGal Dip - Gap Yes MnMo 0.351/0.351 490 570 HAZ
3.8 mm DuraGal Spray - No Gap No MnMo 0.426/0.426 424 535 HAZ
3.8 mm DuraGal Spray - No Gap Yes MnMo 0.426/0.426 440 540 HAZ
8 mm DuraGal Nominal properties - - Unwelded 350 400 -
8 mm DuraGal Measured properties - - Unwelded 399 507 -
8 mm DuraGal Dip - Gap No LW1 0.494/0.494/0.494/0.412 401 486 Parent
8 mm DuraGal Dip - Gap Yes LW1 0.494/0.494/0.494/0.412 406 493 HAZ
8 mm DuraGal Dip - No Gap No LW1 0.498/0.360/0.432/0.360 400 494 HAZ
8 mm DuraGal Spray - No Gap No LW1 0.728/0.494/0.494/0.412 387 474 HAZ
8 mm DuraGal Dip - Gap No MnMo 0.411/0.411/0.411/0.411/0.334 411 496 Parent
8 mm DuraGal Dip - Gap Yes MnMo 0.411/0.411/0.411/0.411/0.334 413 496 Parent
8 mm DuraGal Spray - No Gap No MnMo 0.690/0.585/0.585/0.585 384 473 HAZ
8 mm DuraGal Spray - No Gap Yes MnMo 0.690/0.585/0.585/0.585 386 487 HAZ
Notes: a) For cases of multiple weld runs, the heat input for each welding run is given.
b) Yield stress is the dynamic 0.2% proof stress.
c) Ultimate stress is the dynamic ultimate stress.
d) HAZ was defined as the region of zinc removal caused by welding, rather than by micro or macroscopic observation.
e) Nominal properties obtained from reference [7].
Table 2: Summary of results
7
Figure 12: Failed 8.0 mm specimens Figure 13: Failed 3.8 mm specimens
(The weld is at the centre of each specimen. The extent of zinc removal has been indicated with a
permanent marker on either side of the weld.)
5 DISCUSSION
Several observations can be made from the test results.
There is considerably more variation in the results of the 3.8 mm steel compared to the 8.0 mm steel.
There is a statistically significant drop in yield and ultimate stresses in the welded 3.8 mm steel,
compared to the unwelded material. The change in properties for the 8 mm steel is small.
The measured properties of the unwelded 3.8 mm steel are significantly higher than the nominal
properties. This is very common, but as a result, the strength of the 3.8 mm steel is higher than that
of the commonly used welding wire, Autocraft LW1. It is usual practice to match the strength of the
welding consumable to that of the parent metal. In two instances (3.8 mm, dip method, LW1), fracture
occurred in the weld rather than in the parent metal. For the corresponding case using the spray
method (higher heat input), failure occurred in the HAZ, indicating that the higher heat input had
reduced the strength of the HAZ by a greater amount compared to the dip method.
Welding produces a greater percentage reduction in strength for the 3.8 mm steel, compared to the
8 mm steel. The 3.8 mm steel is more heavily cold-worked to produce its higher nominal strength
compared to the 8 mm steel. Consequently, there is greater scope for strength reduction in the HAZ.
The higher strength electrode (Mn-Mo) produced greater capacity in the 3.8 mm sections compared
to the LW1 electrode despite a similar heat input. The Mn-Mo electrode had an almost negligible
effect on the 8 mm steel, compared to the results of the LW1 electrode.
The higher heat input method of spray transfer compared to dip transfer produces a larger reduction
in yield and ultimate stresses.
There are several instances in which the ultimate strength of the welded 3.8 mm specimens drops
below the yield stress of the parent material. Consequently, a welded connection of this type would
not be able to provide the amount of ductility required for seismic or plastic design applications. The
ultimate strength of the welded 8 mm samples did not fall below the yield stress of the parent material.
8
It should be noted that to utilise the available feedstock most efficiently, BHPSPP use virgin strip with
yield stress fyn = 360 MPa for the 3.8 mm flat bar, and a different strip with yield stress fyn = 300 MPa
for the 8.0 mm flat bar. This is the most likely cause of the 3.8 mm steel exhibiting strength
considerably higher than the nominal properties. It is possible that BHPSPP may change the
feedstock, so that all DuraGal flatbars are produced from the 300 MPa strip. It is possible that flat bar
made from this material will not experience as significant changes in the strength of the HAZ,
compared to the product tested.
This paper has considered the preliminary results of the initial stage of this project. Future
examinations will consider microhardness determination, macro cross section examination, and fillet
welded specimens.
6 SUMMARY
This paper has described tensile tests to examine the effect of various welding procedures on the
mechanical properties of in-line galvanised cold-formed steel. Sections of cold-formed flats were butt
welded together using either the dip transfer method or the spray transfer method using either LW1
or Mn-Mo electrode. In most cases, the tensile specimens failed in the heat affected zone. There was
a small reduction in the yield and ultimate stresses in the welded 8.0 mm steel compared to the
unwelded steel. The 3.8 mm samples displayed a more significant drop in yield and ultimate stresses
when welded, and the drop in strength was greater when the higher heat input spray method was
used. The 3.8 mm steel had a higher nominal strength than the 8.0 mm steel due to more cold
working in the manufacturing process, so it is not unexpected that this steel experienced a greater
drop in strength when welded. Significantly, there were some occasions in which the ultimate strength
of the welded 3.8 mm specimens dropped below the yield stress of the parent material
7 REFERENCES
[1] BHPSPP, (1999), “Structural Cold Formed Hollow Sections and Profiles”, Technical
Information, BHP Structural and Pipeline Products, Mayfield, Newcastle. Australia.
[2] American Welding Society, (1976), Welding Handbook, Vol 1, Fundamentals of Welding, 7th
Edition, (Weisman, C., editor), AWS, Miami, Florida, United States.
[3] Wilkinson T. and Hancock G. J., (1998), “Tests of Knee Joints in Cold-Formed Rectangular
Hollow Sections”, Research Report, No R779, Department of Civil Engineering, The University
of Sydney, Sydney, Australia.
[4] Wilkinson T. and Hancock G. J., (2000), “Tests to examine the plastic behaviour of knee joints
in cold-formed RHS”, Journal of Structural Engineering, American Society of Civil Engineers,
Vol 126, No 3, March 2000, pp 297-305.
[5] Packer, J. A., Wardenier, J., Kurobane, Y., Dutta, D, and Yeomans, N., (1992), Design Guide
for Rectangular Hollow Section (RHS) Joints under Predominantly Static Loading, CIDECT
Design Guide No 3, Verlag TÜV Rheinland GmbH, Köln, Germany.
[6] Standards Australia/Standards New Zealand, (1996), Australian Standard AS/NZS2717.1
Welding - Electrodes - Gas metal arc, Part 1: Ferritic steel electrodes, Standards Australia,
Sydney, Australia.
[7] CIGWELD, (1993), Welding Consumable Guide, Part No. WCGUIDE, Preston, Victoria,
Australia.
9
[8] Standards Australia / Standards New Zealand, (1995), Australian / New Zealand Standard
AS/NZS 1554.1 Structural Steel Welding, Part 1: Welding of Steel Structures, Standards
Australia, Sydney, Australia.
[9] Standards Australia, (1997), Australian Standard AS 2205.1 Methods for destructive testing
of welds in metal; Method 1: General requirements for tests, Standards Australia, Sydney,
Australia.
[10] Standards Australia, (1997), Australian Standard AS 2205.2.1 Methods for destructive testing
of welds in metal; Method 2.1: Transverse butt tensile test, Standards Australia, Sydney,
Australia.
[11] Standards Australia, (1991), Australian Standard AS 1391 Methods for tensile testing of
metals, Standards Australia, Sydney, Australia.
[12] Standards Australia, (1997), Australian Standard AS 2205.2.2 Methods for destructive testing
of welds in metal; Method 2.2: All-weld-metal tensile test, Standards Australia, Sydney,
Australia.
8 NOTATION
The following symbols are used in this paper:
fu Ultimate tensile strength
fun Nominal ultimate tensile strength
fy Yield stress
fyn Nominal yield stress
9 ACKNOWLEDGEMENTS
This paper forms part of a research project entitled “Strength and fracture of steel connections”, jointly
funded under the SPIRT scheme by the Australian Research Council and BHP Structural and Pipeline
Products. The experiments were carried out in the J. W. Roderick Laboratory for Materials and
Structures, Department of Civil Engineering, The University of Sydney. The authors are grateful for
the work of Mr Grant Holgate, in the manufacture of the test specimens, and the advice given by Mr
Paul Grace, WTIA, with respect to welding procedures.
10
