Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02340827 2001-03-14
TITLE
"HIGH TITANIUMIZIRCONIUM FILLER WIRE FOR
ALUMINUM ALLOYS AND METHOD OF WELDING"
RELATED APPLICATIONS
This application is related to copending U.S. Patent Application entitled
Improved
Filler Wire for Aluminum Alloys and Method of Welding, filed on March 15,
1999, and
having Attorney's Docket Number 41992-00236.
TECHNICAL FIELD OF THE INVENTION
This invention relates generally to the field of welding, and more
particularly to
a high titanium/zirconium filler wire for aluminum alloys and method of
welding.
BACKGROUND OF THE INVENTION
Aluminum alloys are used in numerous applications, including car bodies,
storage vessels, and the like. One particular type of aluminum alloy is
aluminum-lithium
alloys. Aluminum-lithium alloys are often used in space vehicles, such as the
external
tank of the Space Shuttle. Welding aluminum alloys, and in particular aluminum-
lithium
alloys, is often difficult due the propensity forweld cracking. Cracks in the
weld reduces
the strength of the weld and can also create a leakage paths in the weld. In
particular,
a poor weld can result in stress concentrators, reduced resistance to low
cycle and high
cycle fatigue, and a reduction in corrosion resistance.
Filler wire is used in the welding process. The chemistry, fabrication, and
welding process of the filler wire can greatly effect the propensity of the
weld to crack.
"Chemistry" refers to the elemental components of the filler wire.
"Fabrication" refers
to the specialized process that the filler wire is fabricated. "Welding
Process" refers to
the particular type of welding process used to produce the weld, such as
variable
-1-
CA 02340827 2001-03-14
plasma arc (VPPA), gas tungsten arc (GTA), gas metal arc (GMA), and soft
plasma arc
(SPA) welding, as well as the specific parameters surrounding the welding
process,
such as preheating and the electrode temperature.
Conventional aluminum alloy filler wires do not sufficiently reduce the number
of
cracks in the weld. As a result, the welded part must often be scrapped or the
weld
repaired. Weld repair is often accomplished by grinding-out the old weld and
then
rewelding the parts. Both scrapping the parts and weld repair are expensive
and time
consuming. Accordingly, the cost and fabrication time of the final product is
increased.
SUMMARY OF THE INVENTION
Accordingly, a need has arisen for an improved filler wire. The present
invention
provides a high titanium/zirconium filler wire for aluminum alloys and method
of welding
that substantially reduces or eliminates problems associated with prior
systems and
methods.
In accordance with one embodiment of the present invention, titanium and
zirconium are added to the filler wire in amounts greater than conventional
filler wires.
The addition of the higher than normal titanium and zirconium reduces crack
susceptibility of aluminum alloy welds, and also enhances the repair
weldability of the
welds, including planished welds. The titanium/zirconium filler wire exceeds
the weld
material properties of conventional weld filler wires.
Other technical advantages will be readily apparent to one skilled in the art
from
the following figures, descriptions, and claims.
_2_
CA 02340827 2001-03-14
BRIEF DESCRIPTION OF THE DRAWINGS
For a more complete understanding of the present invention and the advantages
thereof, reference is now made to the following description taken in
conjunction with the
accompanying drawings, wherein like referenced numerals represent like parts,
in
which:
FIGURE 1 is chart illustrating the chemistry of various filler wires in
accordance
with the present invention;
FIGURE 2 is a chart illustrating VPPA Weld Tensile Properties for the filler
wires
shown in FIGURE 1 in accordance with the present invention;
FIGURE 3 is a chart illustrating RS/Planished Weld Properties for the filler
wires
shown in FIGURE 1 in accordance with the present invention; and
FIGURE 4 is a chart illustrating the data shown in FIGURE 3 in accordance with
the present invention.
DETAILED DESCRIPTION
The present invention is used in the welding of aluminum alloys, and is
particularly suited for fusion welding 2195 aluminum-lithium. The filler wire
can be used
with a number of fusion welding processes, such as variable plasma arc (VPPA),
gas
tungsten arc (GTA), gas metal arc (GMA), and soft plasma arc (SPA) welding.
The
weld filter wire chemistries are an improvement over the conventional filler
wires, such
as 4043 filler wire. For example, compared to 4043, improved weld and repair
weld
mechanical properties are achieved. The additions of titanium and zirconium
significantly improves grain structure refinement in the weld.
-3-
CA 02340827 2001-03-14
During a development program of 2195 Aluminum-Lithium, a 2319 filler wire
chemistry was cast as a control sample for an L8 orthogonal array. The 2319
chemistry
was discovered to contain 0.22 percent zirconium (Zr). This was within the
2319
specification limit of 0.10 to 0.25% Zr; however, it was not the 2319 nominal
value of
0.13% Zr typically used in welding. The 2319 with 0.22% Zr filler wire
chemistry
produced a greatly refined weld grain structure not seen with nominal 2319.
Furthermore, two important observations were made in 0,320-inch-thick 2195-
RT70
variable polarity plasma arc (VPPA) welds. The first observation was weld
fusion line
and crater cracks were significantly reduced at sudden weld stops, or "E-
stops." The
second observation was the elimination of micro-cracks in the first pass of a
two pass
VPPA weld.
During the same time frame, the Space Shuttle's Super Lightweight External
Tank Program (SLWT), was encountering problems repair welding 2195 aluminum-
lithium VPPA welds made with 2319 filler wire. The repair welding involved
"R5" repair
welding and planishing, which was baselined by the SLWT program as the
critical
requirement for filler wire acceptance. The repair welds were produced using
manual
gas tungsten arc (GTA) welding. The R5 repair, which consisted of making five
3.0-
inch-long manual repair welds in the same weld location, simulated a
production repair
having been reworked five times. The R5 repairs were planished to simulate
"oil can"
repairs on the propellant tank caused by the repair weld shrinkage. Then they
were
processed into 1.0-inch-wide straight bar tensile specimens and tested at room
and -
320 degrees F temperatures. Various tensile specimens displayed low tensile
values
at room and -320 degrees F temperature that were below acceptable limits.
CA 02340827 2001-03-14
As a result, development was initiated to provide additional 2319 variants to
improve room temperature and -320 degrees F tensile properties of planished R5
repair
welds. New filler wire chemistries were formulated containing nominal and low
copper
(Cu) with high levels of titanium (Ti) and zirconium (Zr). Four filler wire
chemistries
listed in FIGURE 1 were produced into 1/16-inch-diameter spooled filler wire
by
Reynolds Metals Company. In order to evaluate high and low Cu additions and
the
effect of Mn, chemistries #71106, #71108, and #71109 were selected for VPPA
and
repair weld mechanical property screening. One 0.200-inch-thick x 15.0-inch-
wide x
16.0-inch-long VPPA weld panel was produced with each filler wire chemistry.
On each
panel, one 3.0-inch-long R5 repair weld was performed using direct current
electrode
negative (DCEN) gas tungsten arc (GTA) welding. The repair welds were
performed
with the same filler wire chemistry used in the VPPA weld.
The completed R5 repair welds for mechanical property screening were
inspected, planished, and re-inspected. Three one-inch-wide straight bar
tensile
specimens were taken from VPPA and repair weld areas of each panel and tested
at
room temperature. Two metallographic specimens were taken from both VPPA and
repair weld areas. Because the Mn addition showed no improvements, chemistry
#71109 was eliminated from further testing. Chemistries #71106 and #71108 were
down selected for additional testing. This involved room temperature and -320
degrees
F tensile testing. Two 0.200-inch-thick x 12.0-inch-wide x 24.0-inch-long VPPA
weld
panels were produced with each filler wire chemistry. On each of the panels,
two 3.0-
inch-long R5 repair welds were performed using the DCEN gas tungsten arc (GTA)
welding process. The repairwelds were performed using the same filler wire
chemistry
-5-
CA 02340827 2001-03-14
used in the initial VPPA weld.
The completed R5 repair welds or the down selected filler wire chemistries
were
inspected, planished, and reinspected. Four room temperature and four-320
degrees
F repair weld tensile specimens were tested from each filler wire chemistry.
The tensile
fracture areas from all specimens were processed into metallographic
specimens.
Forthe screening test, 0.200-inch-thick weld test panels were dry machined
from
0.375-inch-thick 2195-RT70 plate produced by RMC, Lot #921 T894A. The 0.200-
inch-
thick weld test panels forthe down selected testing was machined from 0.250-
inch-thick
2195-RT70 plate produced by RMC, Lot #930T649A. Prior to welding, the weld
joint
edges draw filed and adjacent weld joint surfaces scraped. Manual tack welding
of the
weld test panels was performed using DCEN GTA welding. Three autogenous tack
welds were made on the panels, with start and stop tabs welded to the panel
ends.
With 0.063-inch-diameter filler wire, the average weld parameters used in
making the
repairs welds were 85.0 amps, 19.0 volts, and 8.0 ipm travel rate. The
operation -
consisted of performing five simulated repairs in one location of the panel,
alternating
from root side to the face side of the weld area. For example, R1, the first
repair, was
performed on the root side of the VPPA weld, and R2, the second repair, was
performed on the face side. R3, R4 and R5 followed with the same alternating
pattern.
Each repair cycle consisted of filling a 0.100-inch-deep x 0.200-inch-wide x
3.0-inch-
long groove using a two pass GTA weld, which was produced using a dye grinder
with
a carbide cutting wheel. Visual and radiographical inspection was performed
per
MSFC-SPEC-504C on each of the completed R5 repair welds.
-6-
CA 02340827 2001-03-14
After the R5 repair weld operation, the planishing operation was performed.
Planishing was performed in the vertical position using a pneumatm rivet gun
with a
1.86-inch-diameter mushroom head and 2.75-inch-diameter steel bucking bar on
the
opposite side. The repair welds were planished in order to recover
approximately 90
percent of the repair weld shrinkage. Visual, radiographic, and penetrant
inspection
was performed per MSFC-SPEC-504C on each of the planished repair welds.
Radiographic inspection was performed at 45 and 90 degrees from weld panel
surface.
Prior to performing the screening test, repair weld practice panels were run.
These panels displayed toe cracking in the repair welds made with the #71109
and
#71109 filler wire chemistries. The toe cracks ranges from 0.010 to 0.100-
inches-in-
depth and followed the repair welding fusion line, but not necessarily the
VPPA weld
fusion line. For the screening test, the repair welding current was reduced
from 95.0
to 85.0 amps in order to eliminate the possibility of cracking. From the
screening test,
VPPA welds made with chemistries #71106, #71108 and #71109 were visually,
radiographically, and penetrant inspected with acceptable results. The
planished R5
repairwelds made with chemistries #71106 and #71108 were visually,
radiographically,
and penetrant inspected with acceptable results. However, for chemistry
#71109,
radiographic inspection at the R5 level revealed voids in the repair weld,
which was
welder related. The voids were repaired producing a R9 condition. After the
repair, the
#71109 repair was successfully planished and inspected. The averaged VPPA and
planished R5 repair weld tensile data from the screening test is contained in
FIGURE
2. Metallographic examination of the VPPA and planished R5 repair welds showed
pronounced grain refinement when compared to the 2319 filler wire chemistry.
-7-
CA 02340827 2001-03-14
From the down selected chemistries #71106 and #71108, VPPA welds were
visually, radiographically, and dye penetrant inspected with acceptable
results.
Chemistry #71106 had one out of the four R5 repair welds displaying a visual
fusion line
crack after the last cycle of planishing. This was successfully repaired and
planished
in one repair cycle, which made it a R6 repair. Chemistry #71108 had one out
of the
four R5 repair welds displaying a visual fusion line crack after the R5 repair
cycle. This
was repaired in one cycle and planished, which made it a R6 repair. Averaged
room
temperature and -320 degrees F tensile test data is presented in FIGURE 3.
From the weld screening test, all three filler wire chemistries produced
acceptable VPPA and planished R5 repair weld tensile data. Due to the number
of
specimens tested, it is difficult to discuss specific chemistry effects.
However, for VPPA
welds the 6.0% Cu containing filler wire displayed a two ksi higher average
ultimate
strength over the 4.0% Cu filler wire and approximately a four ksi higher
average
ultimate strength over the 4.0% Cu filler wire with 0.3% Mn. The Mn addition
with high
Ti and Zr reduced the ultimate VPPA weld strength: Furthermore, the 4.0% Cu
containing filler wires were susceptible to toe cracks in repair weld practice
panels,
which were run at 95.0 amps of welding current. For the screening and down
selected
tests, the welding current was reduced to 85.0 amps to reduce toe cracking in
the 4.0%
Cu filler wire.
From the available metallographic results, it is revealed that all three
filler wire
chemistries displayed a very fine weld grain structure similar to 2319 filler
wire
containing 0.22% Zr. The fine weld grain structure was produced without V and
Mn
additions. This reveals that high levels of Ti and Zr can cause the fine grain
structure
_g_
CA 02340827 2001-03-14
to form in the weld. From the down selected chemistries, planished R5 repair
weld
tensiles, which were tested at room temperature and -320F, produced acceptable
results, as shown in FIGURE 4. No low tensile values were observed, which had
previously been seen using 2319 and 2319 with 0.22% Zr. Keeping in mind the
quantity of specimens tested, chemistry #71106 and #71108 showed no
significant
differences in repair weld tensile data.
The following conclusions can be made from the filler wire testing. The 2319
variants, containing high levels of Ti and Zr, passed R5 repair welding and
planishing.
There were no low tensile results observed in room temperature and -320F
tensile tests
for the 2319 variants. Due to the least amount of repair weld cracking, the
2319 variant
containing six percent copper performed better in repair welding over the four
percent
copper containing filler wires. The addition of 0.3% Mn to the high Ti/Zr
containing
4.0% Cu filler wire reduced the ultimate strength of VPPA welds.
Although the present invention has been described in several embodiments,
various changes and modifications may be suggested to one skilled in the art.
It is
intended that the present invention encompass such changes and modifications
that
fall within the scope of the appended claims.
_g_
