Weldability of a RS-PM Al-8Fe-2Mo Alloy
Application of high-energy-density welding processes produces
three different types of solidification microstructures
BY K. S A M P A T H A N D W . A. BAESLACK HI
ABSTRACT. The weldability of a dispersion-strengthened Al-8Fe-2Mo (wt-%)
alloy produced via rapid solidificationpowder metallurgy (RS-PM) processing
was investigated. Electron beam (EB) and
pulsed Nd:YAG laser welding conditions were found to have a significant
effect on weld solidification behavior,
microstructural developmen t and mechanical properties. Microstructural
analysis revealed an influence of local
thermal conditions (i.e., temperature
gradient and cooling rate), macroscopic
solid/liquid interface growth rate and the
presence of undissolved/unmelted base
metal dispersoid particles on solidification phenomena in the fusion zone.
Three different types of solidification microstructures, each of w h i c h could be
observed to varying degrees within a single weld fusion zone, were observed.
Superior weld mechanical properties,
including joint efficiencies of 100% in
pulsed Nd:YAG laser welds, were
achieved when high energy density
welding conditions produced increasingly refined microstructures.
Introduction
In recent years, alloy development efforts based on rapid solidification-powder metallurgy (RS-PM) processing techniques have produced a new generation
K. SAMPATH is with Concurrent Technologies Corp., Johnstown, Pa. W. A. BAESLACK
III is with the Department of Welding Engineering, The Ohio State University, Columbus, Ohio.
Originally presented by K. Sampath when a
candidate for the Henri Granjon International
Award at the 73rd Annual AWS Meeting,
held March 22-27, 1992, in Chicago, III.
416-s I A U G U S T 1993
of aluminum alloys for elevated-temperature aerospace applications. These alloys are based on hyper-eutectic Al-Fe
compositions with ternary and quaternary additions of transition and rareearth metals (Refs. 1, 2). These alloy additions exhibit low solid-solubility and
low solid-state diffusivity in alpha aluminum (Refs. 3 and 4) and consequently
tend to form dispersoid particles. In contrast to conventional ingot metallurgy
(IM) processing, rapid solidification processing (RSP) of these novel chemistries
suppresses the formation of primary intermetallic particles (Al 3 Fe type) and instead produces particulates exhibiting
metastable microstructures such as AlAl 6 Fe type micro-eutectic and/or a solute-supersaturated dendritic alpha aluminum. The occurrence of these
metastable microstructures is attributed
to the f o l l o w i n g two characteristics of
RSP techniques: 1) subdivision of the
melt isolates potential nucleants and
produces nucleant-free particulates
which undergo nonequilibrium solidification (Ref. 5), and/or(Ref. 2) the effect
of rapid cooling rate, w h i c h does not
provide sufficient reaction time for a po-
KEY WORDS
Electron Beam Welding
Pulsed Nd.YAG Laser
Laser Welding
Powder Metallurgy (PM)
Rapid Solidification
AI-8Fe-2Mo Alloy
Weld Microstructure
Weld Solidification
tential nucleant to produce the equilibrium microstructure (Ref. 6).
Subsequent thermomechanical processing (TMP) of these rapidly solidified
particulates produces a consolidated
product with a unique microstructure
consisting of submicron-sized dispersoid particles in an extremely finegrained (<1 u.m) alpha aluminum matrix. These microstructural features provide dispersion strengthening (Refs. 2, 4
and 6), and exhibit high specific strength
at elevated temperatures up to 345°C
(650°F). In fact, a comparison of the elevated-temperature
tensile
yield
strengths of a dispersion-strengthened
RS-PM AI-8Fe-2Mo alloy with conventional IM alloys used in aerospace applications (Fig. 1) indicates that the A l Fe based alloy exhibits high specific
strength comparable to Ti-6AI-4V. In
view of their superior properties, these
Al-Fe based alloys produced via RS-PM
processing are considered candidate
materials to replace titanium alloys used
in the fabrication of several aerospace
components such as fan and compressor cases, vanes and blades in gas turbine engines and fins, winglets and
rocket motor cases of missiles (Refs. 6,
7) w h i c h operate at elevated temperatures up to 345°C (650°F). The substitution of RS-PM Al-Fe based alloys is expected to allow significant improvements in the thrust-to-weight ratio and
the performance characteristics of these
aerospace components.
Since weight savings are critical for
advanced aerospace structural applications, joining of these RS-PM Al-Fe based
alloys requires the utilization of w e l d ing processes and/or conditions that can
produce high joint efficiency without an
appreciable increase in weld size or re-
Temperature - °C
100
160
200
Table 1 — Mechanical Properties of RS-PM
AI-8Fe-2Mo Sheet
300
- 1000
Aluminum Alloy With Yield
Strength/Density Ratio Equal
to T1-6AMV
120
TV-6AI-4V
80
Fig. 1 — A comparison
of elevated-temperature tensile yield
strengths of dispersionstrengthened RS-PM
AI-8Fe-2Mo alloy with
conventional IM alloys
for aerospace applications (Ref. 2}.
UTS, Longitudinal(a>
"o Elongation
UTS, Transverse'13'
"b Elongation
800
- 600
510 MPa (74 ksi)
14
376 MPa (54.5 Ksi)
11
CL
(a) Average of 3 values (500, 512, 518)
(b) Average of 3 values (366, 388, 374)
400
40
200
750
50
Temperature - °F
i n f o r c e m e n t . Further, c o n s i d e r i n g the relationships between the unique m i c r o s t r u c t u r e a n d the s u p e r i o r m e c h a n i cal p r o p e r t i e s o f t h e RS-PM base m a t e rial, weldability considerations require
u t i l i z a t i o n o f w e l d i n g processes a n d / o r
conditions w h i c h can "recreate" a
rapidly-solidified microstructure and/or
" r e t a i n " t h e h i g h p e r f o r m a n c e base
m e t a l m i c r o s t r u c t u r e in t h e w e l d z o n e .
A n earlier investigation involving the
m e t a l l o g r a p h i c c h a r a c t e r i z a t i o n of a n
e l e c t r o n b e a m (EB) w e l d in a RS-PM A l 8Fe-1.7Ni alloy s h o w e d that the rapids o l i d i f i c a t i o n a n d c o o l i n g c o n d i t i o n s associated w i t h t h e h i g h e n e r g y d e n s i t y EB
w e l d i n g process offered a s t r o n g p o t e n tial for " r e c r e a t i n g " a rapidly solidified
m i c r o s t r u c t u r e in t h e f u s i o n z o n e (Ref.
8). H o w e v e r , a r e l a t i v e l y h i g h h y d r o g e n
c o n t e n t (>5 m L / 1 0 0 g o f Al) of t h e base
m e t a l resulted in e x c e s s i v e f u s i o n z o n e
porosity and precluded evaluation of
m i c r o s t r u c t u r e / m e c h a n i c a l p r o p e r t y relationships.
T h e present w e l d a b i l i t y i n v e s t i g a t i o n
u t i l i z e d a l o w - h y d r o g e n (<1 m L / 1 O O g
o f Al) RS-PM A I - 8 F e - 2 M o a l l o y in o r d e r
t o e v a l u a t e t h e effects o f w e l d p r o c e s s
c o n d i t i o n s o n the d e v e l o p m e n t o f f u s i o n
zone microstructure and w e l d mechanical properties.
Objectives
A c o m p a r i s o n of t h e r a p i d s o l i d i f i c a tion
conditions
which
produce
m e t a s t a b l e m i c r o s t r u c t u r e s i n RSP p a r ticulates w i t h w e l d fusion z o n e solidific a t i o n c o n d i t i o n s reveals s i g n i f i c a n t differences. These differences arise m a i n l y
from the f o l l o w i n g characteristics of a
fusion w e l d :
1) t h e p r e s e n c e of a r e a d i l y w e t t a b l e
base metal substrate;
2) t h e p r e s e n c e of o x i d e i m p u r i t i e s
w h i c h c a n s e r v e as h e t e r o g e n e o u s n u c l e a t i n g agents in the f u s i o n z o n e ;
3) an a b s e n c e of a n y n u c l e a n t i s o l a t i o n e f f e c t (Ref. 5) in t h e w e l d f u s i o n
zone.
In v i e w o f t h e a b o v e d i f f e r e n c e s , t h e
p r e s e n t i n v e s t i g a t i o n w a s a i m e d at d e -
veloping a fundamental understanding
of the i n f l u e n c e of w e l d i n g process a n d
p a r a m e t e r s o n the s o l i d i f i c a t i o n b e h a v i o r o f r e p r e s e n t a t i v e EB a n d p u l s e d
N d : Y A G laser w e l d s in an R S - P M
A I - 8 F e - 2 M o alloy via detailed m i crostructural characterization and to
correlate this b e h a v i o r w i t h w e l d i n g
conditions and parameters. The pulsed
N d : Y A G laser w e l d i n g process w a s u t i l i z e d in this investigation as this process
p r o v i d e d 1.06 u.m w a v e l e n g t h r a d i a t i o n ,
w h i c h is m o r e e f f i c i e n t l y a b s o r b e d b y
a l u m i n u m a n d its a l l o y s c o m p a r e d t o
1 0 . 6 u.m w a v e l e n g t h C 0 2 laser r a d i a t i o n . From a broader perspective, this
investigation w a s also p u r p o s e d to f a c i l itate a f u n d a m e n t a l u n d e r s t a n d i n g o f m i c r o s t r u c t u r a l d e v e l o p m e n t in t h e w e l d
f u s i o n z o n e in a d v a n c e d s t r u c t u r a l m a t e r i a l s s u c h as p a r t i c u l a t e r e i n f o r c e d
metal-matrix composites ( M M C ) and intermetallic matrix composites (IMC)
w h i c h o b t a i n strengthening f r o m the
p r e s e n c e of h i g h m e l t i n g p o i n t d i s p e r soid particles.
Experimental Procedure
T h e A l - 8 F e - 2 M o a l l o y (actual c h e m istry: 8 . 0 w t - % Fe a n d 2.3 w t - % M o )
studied in this investigation originated
as r a p i d l y - s o l i d i f i e d p o w d e r p r o d u c e d
b y Pratt a n d W h i t n e y ' s r a p i d s o l i d i f i c a t i o n rate (RSR) p r o c e s s . In t h e RSR p r o -
Table 2 — Welding Parameters and Weldment Characteristics of AI-8Fe-2Mo Sheet
T r a n s v e r s e - W e l d Tensile Test
Weld#
Travel Speed
mm/s
Energy Input
|/mm
Threshold Bending
Strain %
EB Welding'11' (weld oriented transverse to the rolling direction)
1
12.7
27.6
2
14.8
23.6
3
16.9
23.6
4
16.9
20.7
5
23.3
19.3
6
19.1
18.4
7
21.2
I6.5
3.3
3.3
4.0
4.0
4.0
4.0
UTS
MPa (Ksi)
EL
%
Failure
Location
Joint
Efficiency %
330' d >(46.5)
1.6
FBR
65
5.6
FBR
73
e
370< > (53.5)
4.0
8
23.3
15.0
5.0
Pulsed Nd:YAG Laser Welding" 3 ' (weld oriented along the rolling direction)
9
2.5
41.3'c)
11.0
374«>(54.2)
11
UBM
100
(a) EB Welds Nos. 1 to 8. acceleration voltage 100 kV, current 3.5 mA except for N o . 3 (4.0 mA) and N o . 5 (4.5 mA).
(b) 10 Hz pulse rate, 5.0 ms pulse duration, 105 W pulse p o w e r (10.5 |/pulse), 2.1 k W peak p o w e r , 102 m m focal length, ^ 1 0 6 W / m m 2 p o w e r density, 16.5 L/min argon shielding gas flow rate.
FBR - fusion boundary region; UBM — unaffected base metal.
(c) Average energy input.
(d) Average ol 4 values (342, 320, 326, 332).
(e) Average of 4 values (368, 370, 376, 366).
(f) Average of 4 values (370. 370, 379. 376).
W E L D I N G RESEARCH S U P P L E M E N T I 417-s
A
B
Fig. 2 — A — Light micrograph; B — TEM micrograph of the as-received 1.27-mm AI-8Fe-2Mo sheet. Arrow in A indicates rolling direction.
cess, a rapidly spinning disk disperses a
homogeneous melt of uniform chemical composition into fine, spherically
symmetric powder particles by centrifugal atomization (Ref. 9). A high velocity
(100 m/s) helium gas jet utilized in conjunction with the rapidly spinning disk
provides convective c o o l i n g under
chemically inert conditions. The inert
atomization conditions minimize surface oxidation and reduce subsequent
hydration of the powder particles,
thereby enabling production of a base
alloy with a minimal hydrogen content.
From a solidification standpoint, the helium shroud provides a steep temperature gradient in the range from 10 3 ° to
10 4 °C/mm across the powder particle
and depending on the powder particle
size generates liquid cooling rates in the
range of 10 4 ° to 10 6 °C/s (Ref. 4). These
rapid solidification conditions produce
a micro-eutectic (Al-AI 6 Fe type) or solute-supersaturated dendritic alpha aluminum structure in the individual powder particulates (Refs. 4, 6). Subsequent
inert compaction of the powder and
TMP (extrusion followed by warm
rolling) produced sheet 1.27 mm in
thickness. Table 1 shows selected mechanical properties of 1.27-mm thick
RS-PM Al-8Fe-2Mo sheet.
Autogenous, full-penetration EB and
pulsed Nd:YAG laser welds were produced in the 1.27-mm thick Al-8Fe-2Mo
sheet. Table 2 shows the specific EB and
pulsed Nd:YAG laser welding parameters. The EB welding conditions typically
employed a power level of 3.5 kW to
4.5 kW. Although the pulsed Nd:YAG
laser welding conditions employed a
lower power level (peak power of 2.1
kW), the sharp focus conditions produced a higher power density (~ 10 6
W / m m 2 ) compared to the EB welding.
A second Nd:YAG laser weld was produced at a higher power level (4.6 kW
418-s I A U G U S T 1993
peak power, 230 W pulse power, or 23
j/pulse) in an overlapping spot mode
(about 4 0 % overlap) to provide enhanced delineation of solidification microstructures.
Following radiographic inspection,
the weld coupons were sectioned and
the transverse sections were mounted in
epoxy, ground to 600 grit emery, and
polished w i t h a 3-u.m diamond c o m pound followed by a final polish using
a colloidal silica suspension. In order to
examine the effects of successive laser
pulses on the development of fusion
zone microstructures, the top section of
the pulsed Nd:YAG laser welds was also
prepared for metallographic characterization. The metallographic samples
were etched using Keller's reagent (2.5
mL H N 0 3 +1.5 mL HCL + 1 mL HF +
95 mL distilled water) for characterization using light microscopy and scanning-electron microscopy (SEM).
Base metal and representative weld
coupons were also characterized using
transmission-electron
microscopy
(TEM). Thin foils for TEM characterization were prepared by a twin-jet electropolishing technique using a solution
of 1 part nitric acid in 3 parts methanol,
at 20 V and at a temperature o f — 2 5 ° C .
TEM characterization mainly involved
amplitude contrast imaging under
bright-field (BF) and dark-field (DF)
imaging conditions and was performed
using a JEOL 200CX microscope operated at 200 kV. The bright-field and
dark-field imaging conditions were employed mainly to estimate the alpha aluminum grain size and the size of the second-phase particles and to recognize the
weld solidification growth morphologies. In addition to the amplitude contrast imaging, microchemical analysis
was carried out in the scanning-transmission-electron microscopy (STEM)
mode using a Tracor-Northern TN2000
energy-dispersive spectrometer (EDS)
system equipped with a beryllium w i n dow detector at a 72-deg take-off angle.
A Cliff-Lorimer data analysis algorithm
based on empirically determined K factors (Ref. 10) was utilized to obtain semiquantitative compositions from EDS
data. These semiquantitative compositions were utilized in conjunction with
the published data on Al-Fe-based a l loys and Al-Fe and A l - M o binary systems in order to elucidate the solidification behavior of the weld fusion zone.
Identification and determination of the
crystal structure of the individual phases
in the fusion zone and determination of
the orientation relationships between
the individual phases were not w i t h i n
the scope of the current investigation.
Following light microscopy examination, Vickers microhardness (DPH, 500g load) traverses were obtained from the
transverse sections of the EB welds and
from the transverse and top sections of
the pulsed Nd:YAG laser weld. In order
to correlate microstructure with weld
mechanical properties a limited mechanical property testing of the w e l d ments was also performed. This included
transverse-weld-oriented tensile testing
of selected weldments (Nos. 3, 5 and 9
in Table 2) and longitudinal-weld-oriented bend testing of sections of all
weldments. Longitudinal-weld-oriented
bend testing was performed with the
weld face in tension and the threshold
bending strain (minimum strain to cause
cracking) was determined using the following relationship:
e = t/(2R + t)
(1)
where: e is the outer fiber strain, t is the
thickness of the test coupon and R is the
radius of the die-block used for bending.
Fig. 3 — EB weld (energy input 15 J/mm)
in AI-8Fe-2Mo sheet:
A — macrostructure, large arrows
indicate fusion boundary, small arrows
indicate HAZ bounding fusion zone;
B — microstructure near the fusion
boundary;
C — microstructure near weld center.
ILgfe»:.#W
c -*-- -
ft *,--S
-
|sraHp|
3|il*0
Swig
" ^
IJp%5
* ^ - - • • ,
-
-•„
-
• •
:.~«*ri
20 lim
Results and Discussion
Base Metal Characterization
Light microscopy examination of the
as-received AI-8Fe-2Mo sheet oriented
parallel to the rolling direction showed
a uniform distribution of fine-sized,
dark-etching dispersoid particles in a
light-etching alpha aluminum matrix —
Fig. 2A. Dark-etching stringers were also
occasionally observed along the rolling
direction. These stringers appeared to
originate from coarse powder particles
w h i c h contained appreciable coarsesized primary intermetallic particles
(Ref. 3). Boundaries between individual
powder particles were not evident indicating the effectiveness of the thermomechanical processing treatment. Consistent with the complete inert RSR and
TMP treatments, the microstructure did
not show significant oxide stringers.
TEM examination of thin foils from
the Al-8Fe-2Mo base alloy revealed a
uniform distribution of submicron-sized
dispersoid particles in a fine-grained
alpha aluminum matrix — Fig. 2B.
Bright-field and dark-field imaging indicated the alpha aluminum grain size to
range from 1 to 2 u.m. Semiquantitative
STEM/EDS analysis of the base metal
alpha aluminum grains indicated appreciable supersaturation of both Fe (~ 0.35
wt-% ) and Mo (~ 0.7 wt-% ) relative to
the equilibrium maximum solid solubility limit of both Fe (0.04 wt-%) and Mo
(0.25 wt-%) in alpha aluminum (Refs.
11, 12). Additionally, TEM observations
indicated the dispersoid particles to typically exhibit either a spherical (polygonal) or an acicular morphology. In view
of the submicron size of these particles,
the specific dispersoid types could not
be identified using selected-area electron diffraction techniques. However,
STEM/EDS compositional analysis indicated that the spherical dispersoids were
enriched in both Fe (~ 1 8 wt-% or 10.25
at. -%) and M o (~ 8.3 w t - % or 2.75 at.
-%). These spherical particles were likely
Al 6 Fe type, w h i c h is a metastable, partially coherent strengthening intermetallic previously reported in RS Al-Fe-Mo
and Al-Fe alloys (Ref. 3). STEM/EDS
analysis of the acicular particles indicated a relative increase in Fe content
(~ 30 w t - % or 1 7.38 at. -%) and a significant decrease in Mo content (~ 1.5
wt-% or 0.52 at. -%). These particles pre-
sumably belonged to the equilibrium
6-Al 3 Fe or A l 1 3 F e 4 type intermetallics
(Refs. 3, 11).
W e l d Structure Characterization
Electron Beam Weldments
Light microscopy examination of EB
welds showed a moderate depth-towidth ratio FZ comprised of a light-etching region near the fusion boundary, and
a relatively uniform, dark-etching region
in the central fusion zone, and a noticeable HAZ outside of the fusion boundary. Figure 3A shows the macrostructure
of EB weld No. 8. SEM examination of
the light-etching fusion boundary region
(FBR) at increased magnification revealed second-phase particles exhibiting an acicular (high aspect-ratio) morphology, randomly oriented in a matrix
of relatively coarse-sized dendritic alpha
aluminum (Fig. 3B). Occasional, fine
spherical particles were also observed
in this region. SEM examination of the
dark-etching region w i t h i n central fusion
zone
showed
fine-sized
second-phase particles exhibiting an
equiaxed morphology, uniformly dis-
W E L D I N G RESEARCH SUPPLEMENT I 419-s
Fig. 4—TEM micrographs of EB weld (energy
input 15 j/mm) in Al-8Fe-2Mo. A — Acicular
AljFe type intermetallic panicles and dendritic
alpha aluminum matrix; B — interdendritic AlAlgFe type metastable lamellar eutectic in the
light-etching fusion boundary region; C —
equiaxed AI^Fe type intermetallic particles and
solute-rich particles in alpha aluminum intersticies in the dark-etching central fusion zone.
persed in a fine-sized dendritic alpha
aluminum matrix (Fig. 3C). In addition
to the difference in the morphology of
the second-phase particles, the weld
center showed a significant increase in
the population of second-phase particles vs. the fusion boundary region.
TEM examination of the light-etching
FBR confirmed the presence of acicular
particles in a matrix of coarse-sized dendritic alpha aluminum — Fig. 4A. The
size of the alpha aluminum dendrites
was approximately 2 jun. STEM/EDS
analysis of the alpha aluminum dendrites revealed appreciable supersaturation at =0.42 wt-% Fe (0.20 at. -%) and
=0.76 wt-% Mo (0.21 at. -%) relative to
the equilibrium solid solubility limits of
both Fe and M o in alpha aluminum
(Refs. 11,12). STEM/EDS analysis of the
acicular particles showed s29.2 w t - %
Fe (16.84 at. -%) and =1.6 wt-% M o
(0.54 at. - % ) , indicating these particles
were likely Al 3 Fe type intermetallic.
TEM examination also revealed the occasional presence of fine-sized spherical particles in this region. STEM/EDS
analysis of these spherical particles indicated that they were likely Al 6 Fe type.
TEM examination of the alpha aluminum
dendrite intersticies showed a fine-sized
lamellar eutectic microstructure — Fig.
4B. STEM/EDS analysis of the interlamellar particles showed a typical composition of 219.4 w t - % Fe (10.94 at. %) and s6 wt-% Mo (1.97 at. -%), indicating an Al 6 Fe type metastable intermetallic.
420-s I A U G U S T 1993
TEM examination of the dark-etching
central fusion zone showed coarse-sized
equiaxed-dendritic
particles in a
fine-grained dendritic alpha aluminum
matrix, and solute-rich particles at the
alpha dendrite intersticies — Fig. 4C.
The side arms of the coarse-sized intermetallic particles were about 2 (im in
length, and the size of the alpha aluminum cells ranged from about 0.5 to 1
um. STEM/EDS analysis showed the
well-developed equiaxed particles to
contain typically =26.7 w t - % Fe (15.3
at. -%) and =.2.9 wt-% (0.97 at. -%) M o
indicating that the particles were likely
Al 3 Fe type. TEM examination using BF
and DF imaging did not indicate any orientation relationship between the
equiaxed intermetallic particles and the
dendritic alpha aluminum matrix. The
size of the solute-rich particles at the
alpha aluminum dendrite intersticies
was in the range of 0.1 to 0.2 (im.
Correlation of the alpha aluminum
dendrite spacing in the central weld fusion zone w i t h previous dendrite arm
spacing/cooling rate relationships for
aluminum alloys (Refs. 13, 14) indicated
a weld cooling rate between 10 3 and
10 4c C/s. It is interesting to note that although the use of such moderately rapid
fusion zone cooling rates facilitated
marginal substitution of M o atoms for
Fe atoms within the intermetallic particles, the growth of equilibrium Al 3 Fe
type intermetallic particles could not be
suppressed within the bulk fusion zone.
Further,
although
the
RS-PM
Al-8Fe-2Mo base metal contained Al 6 Fe
type dispersoid particles, the EB weld
fusion zone did not show significant evidence for the presence of AI 6 Fe type intermetallic particles.
The metallographic examination of
the other EB welds (Nos. 1-7 in Table
2) did not show appreciable differences
in either the type or the morphology of
the microstructural constituents in the
fusion zone compared to weld No. 8,
although the proportion of light-etching
to dark-etching microstructural regions,
and the width of the HAZ generally increased with an increase in weld energy
input. It is of importance to note that the
coarse fusion boundary structure exhibited by EB welds in this Al-8Fe-2Mo
alloy contrasted previous results on the
FZ microstructure of a high depthto-width ratio EB weld in 25.4-mm ( 1 in.) thick Al-8Fe-1.7Ni alloy (Ref. 8),
which found epitaxial solidification to a
Al-Al 6 Fe type micro-eutectic structure,
and a gradual increase in the size and
quantity of equilibrium AI 3 Fe type intermetallics toward the weld center. The
occurrence of the micro-eutectic structure in the AI-8Fe-1.7Ni alloy was attributed to the extremely steep temperature gradient and rapid cooling rates
experienced near the fusion line for the
high depth-to-width ratio EB weld (Ref.
8). The absence of such microstructures
in the FZ of EB welds in the AI-8Fe-2Mo
sheet is attributed to the effects of shallow temperature gradient (as indicated
by the wide widths of the HAZ in these
Fig. 5 — Macrostructure of pulsed Nd.YAG laser weld in Al-8Fe-2Mo sheet: A • • transverse section; B — plan view. Arrows in A and B indicate
curvilinear bands bounding successive melt zones.
weldments) and slower cooling rate
commonly occurring in thin section
weldments.
Pulsed Nd:YAC Laser Welds
Light microscopy examination of
transverse and plan sections of the
pulsed Nd:YAG laser w e l d (No. 9 in
Table 2) showed a moderate depthto-width ratio FZ exhibiting an appreciable variation in etching response, and
a negligible HAZ (relative to the EB weld
shown in Fig. 3A) bounding the fusion
zone (Fig. 5A and 5B). The minimal
width of the base metal HAZ observed
in the pulsed Nd:YAG laser weld was in
agreement with the steep temperature
gradient produced by the pulsed
Nd:YAG laser welding conditions. A l though a light-etching microstructure
containing "swirls" tended to dominate
the fusion zone, relatively dark-etching
microstructures were observed as
"curvilinear bands" bounding successive melt zones and in the HAZ with the
base metal bounding the fusion zone.
Examination of the light-etching FZ
structures at increased magnification revealed extremely fine particles within
the alpha aluminum matrix— Fig. 6A.
The presence of swirls showing marginal
variations in the distributions of these
particles appeared to be associated with
the convective fluid f l o w experienced
in the melt prior to weld solidification.
The dark-etching curvilinear bands observed in the boundaries between successive melt zones exhibited a smooth
interface with the original melt zone and
an irregular interface bounding the subsequent melt zone — Fig. 6B. The irregular interface frequently exhibited a
"cusped" appearance, w h i c h was attributed to the nature of the fluid flow
w i t h i n the molten weld pool. Analysis
of this dark-etching structure at increased magnification suggested that it
was comprised of a HAZ with the original melt zone (smooth interface) and a
FBR with the subsequent melt zone (irregular interface). In contrast with the
base metal HAZ, the FBR showed
coarse-sized primary intermetallic particles in a coarser-dendritic alpha aluminum matrix. In this region, the particles appeared to have coarsened in a
stagnant region of liquid alpha aluminum. Additionally, several dark-etching equiaxed particles were also observed in the central FZ.
Evidence of appreciable fluid f l o w
and its influence on dispersoid distribution in the fusion zone was even more
evident in the weld produced using limited overlap between individual spot
welds — Fig. 7A-C. TEM examination
of the weld shown in Fig. 7 provided increased resolution of the weld FZ and
•Js5V.-r-.k.--Jfi
•
50 (im
A
B
Fig. 6 —Microstructure of pulsed Nd:YAG laser weld (No. 9 in Table 2) in Al-8Fe-2Mo sheet. A — Fine-sized particles in the light-etching regions of the bulk fusion zone; B — HAZ in the base metal bounding the fusion zone (arrow). Curvilinear bands in A and B show HAZ in the preceding melt zone (smooth interface) and FBR in the subsequent melt zone (irregular interface).
W E L D I N G RESEARCH SUPPLEMENT I 421-s
Fig. 7 — Microstructure of pulsed Nd:YAG laser
weld (overlapping spot welds) in Al-8Fe-2Mo
sheet. A — Transverse section; B — fop surface; C
— transverse section showing top corner of fusion
zone. Small arrows indicate HAZ bounding successive melt zones. Dark-etching swirls
correspond to dispersoid coarsened regions.
HAZ microstructures. TEM observations
of the light-etching FBR (near the outer
periphery of the weld with the base
metal) using BF and DF imaging conditions revealed fine-sized spherical particles and epitaxial growth of c o l u m nar-dendritic alpha aluminum from base
metal alpha aluminum grains — Fig. 8A.
The size of the individual alpha aluminum dendrites was about 2 | i m .
STEM/EDS analysis found the spherical
particles within the columnar alpha alu-
minum dendrites to typically contain
17.1 w t - % Fe (9.72 at.-%) and 8.7 wt% Mo (2.87 at.-%), indicating these were
likely AI6Fe type intermetallics. The fine
size and spherical morphology of these
intermetallics and their AI 6 Fe type composition suggested their origin from the
base metal.
Figure 8B shows the light-etching region of the fusion zone, which consists
of spherical particles in a matrix of extremely fine dendritic alpha, or possibly
a microeutectic. As near the fusion
boundary, the AlgFe composition of
these particles, and their size and morphology suggested their possible origin
from the base metal.
The dark-etching central FZ regions
exhibited relatively coarse-sized particles which nucleated grains of dendritic
alpha aluminum — Fig. 8C. STEM/EDS
analysis of the coarse-sized particles revealed the composition as s23.8 w t - %
Fe (14.1 at.-%) and s8.25 wt-% M o (2.8
Fig. 8 — TEM micrographs of pulsed Nd:YAG laser weld (overlapping spot welds) in AI-8Fe-2Mo sheet: A — FBR adjacent to base metal;
FZ (light-etching regions in Fig. 7); (continued on next page)
422-s I A U G U S T 1993
A
B
Fig. 9 — TEM micrographs of pulsed Nd:YAG laser weld (overlapping spot welds) in Al-8Fe-2Mo sheet. A — HAZ within FZ of preceding melt
zone; B — HAZ in base metal. Arrow in A indicates spherical intermetallic particles. Arrows in B indicate the boundary between HAZ and unaffected base metal.
at.-%). These coarse-sized particles were
presumably either AI 3 Fe or Al 1 3 Fe 4 type
intermetallics. Nearer to the center of
the FZ, where the temperature gradient
became shallower, these (intermetallic)
particles exhibited equiaxed growth.
STEM/EDS analysis of the alpha aluminum sheath ("halo") around these
coarse particles indicated significant increases in solid-solubility of both Fe
(-4.67 wt-%) and M o (-2.1 wt-%). Interestingly, a comparison of the composition of the overall matrix structure ( =7
w t - % Fe and s2.2 w t - % Mo) w i t h that
of the alpha aluminum sheath suggested
possible alpha aluminum
growth
through solute partitioning. The compositional analysis of the overall matrix
structure was well beyond the eutectic
composition of 1.8 wt-% Fe (Ref. 11) in
binary Al-Fe systems (albeit M o additions influence this composition to a limited extent). The solidification of this hypereutectic composition to a hypo-eu-
tectic microstructure consisting of primary alpha with s4.67 w t - % Fe (vs. a
maximum equilibrium solid-solubility
of 0.04 wt-% Fe (Ref. 11) at the eutectic
temperature) and =2.1 wt-% Mo was attributed to the rapid cooling and solidification rates experienced w i t h i n the
pulsed laser w e l d . Indeed, a comparison of the dendrite arm spacings with
previous relationships between dendrite
arm spacing and cooling rate in aluminum alloys (Refs. 13, 14) indicated a
Nd:YAG laser weld cooling rate in excess of 10 5 °C/s through the solidification temperature range. Compared with
the EB weld cooling rate, the laser weld
cooling rate was at least an order of magnitude higher.
TEM analysis of the "cusped" region
between successive laser pulses revealed both the irregular shape of the
particles and the coarse size of the dendritic alpha aluminum matrix — Fig. 8D.
STEM/EDS analysis indicated that the ir-
regular particles typically contained
27.91 wt-% Fe (1 6.1 at.-%) and 2.63 wt% Mo (0.88 at.-%), and suggested that
these particles were the equilibrium
Al 3 Fe or Al 1 3 Fe 4 type intermetallics. In
comparison with the composition of the
alpha aluminum matrix in the light-etching regions, the coarser as-solidified
alpha aluminum dendrites in the FBR
were considerably leaner in both Fe
(-0.6 wt-%) and Mo (-1.6 wt-%), suggesting that the coarsening of the equilibrium dispersoid particles occurred in
molten alpha aluminum and depleted
the surrounding melt of both Fe and Mo.
TEM analysis of the H A Z within the
preceding melt zone revealed the formation of submicron-sized acicular particles in an alpha aluminum matrix —
Fig. 9A. In comparison with the spherical particles observed in the light-etching regions of the FZ, STEM/EDS analysis showed these acicular particles to exhibit a higher Fe/Mo ratio, suggesting an
Fig 8 (continued): C — center of FZ (dark-etching in Fig. 7);D — FBR between successive laser melt zones. Arrow in A indicates columnar-dendritic growth direction.
W E L D I N G RESEARCH SUPPLEMENT I 423-s
i^'Vju***^ "
HSS
C
D
F/g. 70 — Tf/VI micrographs of pulsed Nd:YAG laser weld (No. 9 in Table 2) in Al-8Fe-2Mo steel. A — FBR adjacent to the base metal; B — center of FZ; C — cusped region between successive melt zones; D — HAZ in the preceding melt zone. Arrows in B indicate spherical intermetallic
particles which experienced equiaxed growth. Large arrows in B indicate HAZ in the preceding melt zone. Arrows in D indicate spherical intermetallic particles.
Al 3 Fe type. During the laser weld thermal cycle, the extremely fine intermetallics or solute-rich particles formed
by elemental segregation at the
last-to-solidify dendrite intersticies apparently underwent preferential dissolution in solid state and thereby provided
solute flux for the formation of these acicular particles. The supersaturated as-solidified alpha aluminum matrix (-3 wt% Fe and s1.9 wt-% Mo) might have
also provided additional solute flux for
the formation/growth of these intermetallic particles. Note that despite the
near-complete absence of the as-solidified dendritic structure, spherical intermetallics w h i c h were previously observed in the as-solidified microstructure are also present in this region.
TEM analysis of the HAZ microstructure in the base metal directly adjacent
to the FZ showed a slight reduction in
dispersoid particle density vs. the base
material suggesting a marginal coarsening of the dispersoid particles arising
from the rapid weld thermal cycle —
Fig. 9B.
TEM examination of the pulsed
424-s I A U G U S T 1993
Nd:YAG laser weld (No. 9 in Table 2)
showed microstructural transitions (Fig.
10) similar to those observed in the weld
produced using limited overlap between
individual spot welds. Consistent with
the lower weld energy input (1 0.5 J/pulse
vs. 23 J/pulse for the overlapping spot
weld) this weld exhibited a refined m i crostructure. Typically, the size of the
alpha aluminum dendrites ranged from
0.5 to 2 u.m. Although not reported here,
the STEM/EDS data obtained from several regions of this w e l d showed only
marginal differences from the STEM/EDS
data obtained from similar locations in
the overlapping spot weld. These differences were consistent with the differences in the levels of overall weld energy input between these two welds, and
could be readily attributed to the effects
of the weld thermal cycle and local weld
cooling rate.
Weld Mechanical Properties
Table 2 shows the results of the weld
mechanical property tests. Consistent
with the microstructural analysis, trans-
verse-weld tensile testing of EB weld
coupons showed lower weld tensile
strengths relative to the base metal (330
and 370 MPa vs. 51 0 MPa) and preferential failure through the coarse fusion
boundary region in the weld fusion
zone. Microhardness traverses across
the welds showed a relatively lower
hardness in the fusion boundary region
compared to the base metal (about 110
DPH vs. 120 DPH). The hardness of the
central fusion zone was generally higher
than that of the base metal and increased
(over 1 30 DPH) with decreasing energy
input. Consistent w i t h the overall m i crostructural refinement observed in the
low energy EB welds, and change in the
proportions of light-and dark-etching
microstructures, mechanical property
testing showed an appreciable increase
in the transverse weld tensile strength,
% elongation and joint efficiency with
decreasing weld energy input. However,
the longitudinal weld bend testing of the
EB weld test coupons did not reveal any
change in the threshold bending strain
to crack initiation. This behavior was attributed
to
the
occurrence
of
coarse-sized intermetallic particles in
the fusion boundary region. These particles appeared to promote strain localization and preferential sites for crack
initiation.
Evaluation of the mechanical properties of the pulsed Nd:YAG laser weld
(No. 9 in Table 2) showed excellent correlation with the metallographic analysis. Transverse-weld tensile testing found
fracture to occur in the unaffected base
metal, indicating 100% joint efficiency
and a transverse-weld tensile strength of
374 MPa (54.2 ksi). Microhardness traverses across the weld showed hardnesses in the light-etching fusion zone
regions (220 DPH) and pulse overlap region between successive melt zones
(about 1 60 DPH) to exceed that of the
base metal (1 20 DPH). The hardness of
the fusion boundary region with the base
metal was also generally higher than that
of the base metal (about 180 DPH vs.
120 DPH). Although the base metal HAZ
did exhibit a hardness slightly below that
of the base metal (11 5 DPH), its resistance to fracture during tensile testing
likely resulted from its narrow width and
associated constraint effects. Longitudinal-weld bend ductility testing revealed
fracture initiation and propagation to be
associated with the pulse overlap region
between the successive melt zones (Fig.
11) and a threshold bending strain only
marginally below that of the base metal
( 1 1 % vs. 14%). It is interesting to note
that excepting for differences in the morphology and size of the intermetallic particles and the alpha aluminum matrix,
the microstructure of the pulse overlap
region between the successive melt
zones in the pulsed Nd:YAG laser weld
was very similar to the fusion boundary
region in the EB welds. This similarity
in microstructural constituents appeared
to explain the preferential crack initiation in these regions during longitudinal
bend testing. The wide differences observed in the threshold bending strain
between the EB and the pulsed Nd:YAG
laser weld were consistent with the wide
variations in the size of these microstructural constituents.
Weld Solidification Behavior
Microstructural characterization described above has shown that the solidification microstructure observed
within the fusion zone of EB and pulsed
Nd:YAG laser welds in Al-8Fe-2Mo vary
widely with respect to alpha aluminum
grain morphology and dendrite size, the
level of alloying element supersaturation of alpha aluminum dendrites, and
the type, size, morphology and population density of second-phase intermetallic particles. Despite such apparently
wide variations, the observed weld soli-
Fig. 11 — Plan view
of longitudinal-weldoriented bend ductility test specimen in
AI-8Fe-2Mo showing
fracture propagation
through the laser
weld pulse overlap
region between successive melt zones.
Arrow indicates
bending direction.
dification structures can be broadly classified into three general types:
Type A —Al 3 Fe type intermetallic
particles in a matrix of dendritic alpha
aluminum with interdendritic lamellar
or divorced eutectic;
Type B — Al 3 Fe type intermetallic
particles which act as growth-centers for
dendritic alpha aluminum w i t h interdendritic divorced eutectic;
Type C — Al 6 Fe type intermetallic
particles "entrapped" in a matrix of dendritic alpha aluminum w i t h interdendritic divorced eutectic.
In general, the local mode of fusion
zone solidification and resulting m i crostructure type are dependent on three
principal solidification parameters viz.,
temperature gradient (G L ), solidification
velocity (R) and degree of undercooling
(AT). In this context, the observed variations in the morphology of the primary
Al 3 Fe type intermetallic particles in the
EB weld could be readily attributed to
variations in the local temperature gradient obtained in different regions of the
weld. Similarly, the observed variations
in the size of the alpha aluminum dendrites and the intermetallic particles
could be attributed to the variations in
the local cooling rate (which is a product of temperature gradient and solidification velocity) with higher cooling rate
in pulsed Nd:YAG weld generally producing fine-sized microstructural constituents. Further, the occurrence of
lamellar metastable eutectic microstructure in the FBR of the EB w e l d , and solute-rich particles {i.e., a divorced eutectic) in the dendrite intersticies of
alpha aluminum in EB and laser welds
could be attributed to the the local degree of undercooling and the relative
size of the interdendritic regions during
the terminal stages of weld solidification
at these locations (Ref. 1 5).
The observed variations in the type
and population of intermetallic particles
within the weld fusion zone appeared
to be related to the peak temperature
distribution during fusion welding. Considering that the RS-PM Al-8Fe-2Mo
base metal contained both Al 3 Fe and
Al 6 Fe type intermetallic particles in an
alpha aluminum matrix, the wide difference in temperature between the melting point of the Al 3 Fe type intermetallic
particles (~1150°C) and the liquidus
temperature of the base metal (~ 900°C)
likely promoted the occurrence of unmelted or undissolved base metal dispersoid particles in an essentially alpha
aluminum melt during the on-heating
weld thermal cycle. As shown in Fig. 5A
and 5B, fluid flow w i t h i n the pulsed
Nd:YAG weld appeared to control the
movement of these dispersoids through
the " h i g h " and " l o w " temperature regions in the weld pool and thereby influence their final size and distribution.
These unmelted/undissolved base metal
dispersoid particles likely underwent
acicular or equiaxed growth during subsequent melt cooling to a degree depending on local thermal condition and
weld cooling rate. Under certain conditions of local melt undercooling, these
particles likely served as heterogeneous
nuclei and thereby promoted separate
solidification {i.e., growth) events.
The above considerations show that
complete melt homogeneity (characterized by the total absence of unmelted
or undissolved_ base metal dispersoid
particles) was not obtained within the
bulk fusion zone of EB and pulsed
Nd:YAG laser welds. Although the origins of unmelted and undissolved base
metal dispersoid particles differ from a
physical standpoint, from a metallographic standpoint these particles can
not be adequately distinguished. Henceforth, these particles are referred to as
undissolved particles.
W E L D I N G RESEARCH SUPPLEMENT I 425-s
Type A Weld Solidification
Based on the above observations, the
formation of the light-etching FBR in the
EB weld (Fig. 3B) may be considered to
occur in three stages. Initially, the undissolved base metal dispersoid particles
grow in a relatively stagnant liquid region into randomly-oriented, acicular
primary intermetallic particles and promote solute-depletion in the remaining
melt. Solute depletion in the melt promotes the growth of columnar-dendritic
alpha aluminum, which occurs epitaxially from base metal alpha. Solute partitioning accompanying alpha a l u minum growth promotes a solute-rich
melt at the interdendritic regions which
undergoes terminal solidification to a
metastable Al-AI 6 Fe type lamellar eutectic at the alpha aluminum dendrite
intersticies. Since the alpha aluminum
solidification initiates from the fusion
boundary at relatively low solidification
velocities, the local solidification conditions promote appreciable structural
coarsening and the formation of relatively coarse alpha aluminum dendrites.
The relatively coarse width of the dendrite intersticies at these locations likely
facilitated the formation of a regular
lamellar eutectic structure (Ref. 15). It is
important to note that morphologically
identical microstructures are commonly
observed throughout the fusion zone of
GTA weldments (although on a coarser
scale) produced in these alloy types,
which are characterized by appreciably
shallower temperature gradients and
slower growth rates vs. electron beam
welds.
The formation of the dark-etching microstructure in the central fusion zone
of the EB weld (Fig. 3C) also appeared
to occur by the above sequence. The
shallow temperature gradient in the central fusion zone ahead of the trailing
edge appeared to promote the equiaxed
growth of undissolved base metal dispersoid particles or possibly promote the
nucleation of these particles from the
liquid. This latter mechanism is suggested by the dispersoid density, uniformity and AI 3 Fe stoichiometry. Nucleation and/or growth of the intermetallic
was followed by the non-equilibrium
single phase solidification (growth) of
the
solute-depleted
melt
into
cellular-dendritic metastable (supersaturated) alpha aluminum, and terminal
solidification of the last-to-freeze solute-enriched liquid into fine-sized solute-rich particles at the alpha aluminum
dendrite boundaries. In comparison with
the relatively coarse-sized alpha aluminum dendrites in the FBR, the formation of fine-sized cellular-dendritic
alpha aluminum growth structure in the
central fusion zone of the EB weld may
426-s I A U G U S T 1993
be attributed to relatively higher values
of weld cooling rates (G L x R) obtainable
at this location. The formation of
fine-sized particles at the interdendritic
regions via a divorced eutectic reaction
instead of the regular Al-AI 6 Fe type
lamellar eutectic microstructure is attributed to the relatively fine size of the
dendrite intersticies compared with the
interlamellar spacing of the regular eutectic (Ref. 15). Interestingly, the m i crostructural development in the cusped
region between successive melt zones
in the pulsed Nd:YAG laser weld (Figs.
8D and 10C) can also be described using
the above solidification sequence.
Type B Weld Solidification
The equiaxed morphology and the
relatively coarse size of the dendritic
alpha aluminum grains in the central fusion zone of the pulsed Nd:YAG laser
welds (Figs. 8C and 10B) suggest that
these grains originated from separate solidification events (Ref. 16) independent
of the epitaxial growth from the fusion
boundary. These microstructures exhibit
supersaturated alpha aluminum dendrites in conjunction with AI 3 Fe type intermetallic particles which indicate considerable undercooling subsequent to
the initial growth of these primary intermetallic particles. Further, the formation
of metastable alpha aluminum growth
structures suggest that although the primary intermetallic particles are characterized by poor nucleation potency for
equilibrium alpha aluminum (Ref. 1 7),
these particles may exhibit a higher nucleation potency (catalytic nucleation)
for the metastable (solute supersaturated) alpha aluminum. Interestingly, a
lattice disregistry of 10% or less between
the potential nucleant and alpha aluminum is considered to promote the catalytic nucleation of alpha aluminum
(Ref. 1 7). Since supersaturation of Fe and
M o in aluminum shows a negative deviation from Vegard's law, and reduces
the lattice parameter of alpha aluminum
(Ref. 18), improved lattice-registry between the primary intermetallic particles and metastable alpha aluminum
and the occurrence of shallow temperature gradients in the central fusion zone
may be considered responsible for the
occurrence of the equiaxed-dendritic
alpha aluminum growth structure at this
location. The metastable alpha aluminum growth structure also indicates
that the composition of the melt surrounding the equiaxed intermetallic particles was essentially hypo-eutectic relative to the metastable eutectic reaction.
In such an instance, it can be easily recognized that the spherically symmetrical undissolved Al 6 Fe type dispersoid
particles present in a hypo-eutectic melt
can not undergo growth. W i t h the
growth of the alpha aluminum, the composition of the remaining melt in the
dendrite intersticies is enriched with Fe
and M o and it ultimately reaches the
metastable eutectic composition and undergoes terminal (metastable) eutectic
solidification. In view of the fine size of
the interdendritic region, solute-rich particles are formed in this region instead
of a regular lamellar eutectic (Ref. 15).
It is interesting to note that the Type
B solidification mode is occasionally observed in the dark-etching region of the
EB welds. However, DF imaging of the
alpha-aluminum grains using TEM indicated that solidification of this region
principally occurred by the Type A
mode.
Type C Weld Solidification
The light-etching fusion boundary region in the pulsed Nd:YAG laser weld
(Figs. 8A and 10A) appeared to originate
by epitaxial solidification from base
metal alpha in the fusion boundary.
These epitaxially grown alpha-aluminum grains effectively entrapped the
undissolved dispersoid particles as solidification proceeded along a relatively
steep temperature gradient across the
advancing solid-liquid interface. The occurrence of steep temperature gradients
in this region promoted solidification of
the melt to a columnar-dendritic substructure with fine intermetallics or solute rich particles partitioned to the
last-to-solidify dendrite intersticies. The
high fusion zone cooling rate (in excess
of 10 5 °C/s) did not provide adequate
time for the growth of the undissolved
particles. Consequently, the melt undercooled to a temperature below the
metastable eutectic temperature. At this
level of melt undercooling, solidification initiated via epitaxial growth of
columnar-dendritic
(supersaturated)
alpha aluminum from the base metal
alpha. Coincident with the growth of the
metastable alpha aluminum, the remaining melt is enriched w i t h Fe and Mo
which underwent final solidification to
fine-sized particles due to the relatively
fine size of the interdendritic regions
(Ref. 15). This level of undercooling apparently promoted a high solidification
velocity of the solid/liquid interface
which in turn entrapped the undissolved
globular intermetallic particles at the advancing growth front.
Although epitaxial solidification generally indicates local wetting of the fusion boundary and minimal melt undercooling, the occurrence of epitaxially
grown metastable Al-Al 6 Fe type c o u pled-eutectic microstructures in the f u sion boundary has been previously reported in EB weld produced in RS-PM
Al-8Fe-1.7Ni alloy (Ref. 8). In fact, Fig.
8B shows a possible occurrence of A l AI 6 Fe type micro-eutectic structure in
the pulsed Nd:YAG laser weld in the
RS-PM Al-8Fe-2Mo alloy as well. These
results suggest that despite the presence
of a readily wettable substrate,
metastable growth structures in the weld
fusion zone can likely occur from the
combined effects of a minimal nucleation barrier for the metastable phase, a
steep temperature gradient w h i c h restricts the growth of the stable phase,
and perhaps a minimal difference in the
level of undercooling required to allow
the growth of metastable structures. It is
interesting to note that in hyper-eutectic Al-Fe melts while the nucleation of
Al 3 Fe occurs at less than 8°C undercooling, the nucleation of Al 6 Fe is favored
at undercoolings greater than 10°C (Ref.
11).
Summary
In summary, during fusion welding
of AI-8Fe-2Mo alloy peak temperature
distributions near the fusion boundary
are below the equilibrium liquidus temperature. These peak temperature distributions promote the presence of undissolved base metal dispersoid particles
within the bulk fusion zone. Subsequent
weld solidification within the fusion
zone occurs as one of three different
types depending on local thermal conditions, principally cooling rate and temperature gradient which are dependent
on welding process condition and parameters. These different types of fusion
zone solidification promote the formation of different solidification microstructures and lead to significant microstructural transitions within the bulk
fusion zone.
Conclusions
1) Fusion zone microstructures of EB
and pulsed Nd:YAG laser welds in
Al-8Fe-2Mo exhibit local variations in
the size, morphology, the level of Fe and
M o supersaturation in alpha aluminum
dendrites and appreciable differences in
the type, size, morphology and population of intermetallic particles.
2) Analysis of the solidification behavior of these fusion welds determined
that the above variations in fusion zone
microstructures can be strongly influenced by the presence of undissolved
base metal dispersoid particles and from
the effects of local variations in thermal
conditions within the fusion zone. Solidification of the "fusion zone" was observed to occur in one of three different
modes.
3) Weld mechanical properties testing showed that despite the occurrence
of fracture in the dispersoid-coarsened
fusion boundary region, electron-beam
welds exhibited high joint efficiencies
and good ductility. The absence of a dispersoid-coarsened region adjacent to
the fusion boundary in the pulsed
Nd:YAG laser welds and negligible dispersoid coarsening in a narrow HAZ resulted in 100% joint efficiency.
A cknowledgmen ts
The authors express their deep appreciation to Andrew Crowson and Edward
Chen of the U.S. Army Research Office
for financial support, John Lippold, formerly of Sandia National Laboratories,
Livermore, Calif., for producing the electron beam welds, Tom Lienert formerly
of Sandia National Laboratories, Albuquerque, N.Mex., for producing the laser
welds, H. O. C o l i j n , Central Electron
Optics Facility, The Ohio State University, for his considerable assistance with
the semiquantitative STEM/EDS analysis
and K. H. Hou, Department of Welding
Engineering,, The Ohio State University,
for his assistance with the printing of
TEM negatives.
References
1. Langenbeck, S. L., Griffith, W. M.,
Hildeman, G. J., and Simon, ). W. 1986. Development of dispersion-strengthened aluminum alloys. Rapidly Solidified Aluminum
Alloys, STP 890, eds. M. E. Fine and E. A.
Starke, Jr. ASTM, pp. 410-422.
2. Bourdeau, R. G., Adam, C. M., and van
Reuth, E. 1 982. Application of rapid solidification to gas turbine engine. Proc. International Conf. Rapidly Quenched Metals,
Sendai, 1981, eds. T. Masumoto and K.
Suzuki, The Japan Inst. Metals, pp. 155-158.
3. Adam, C. M., and Bourdeau, R. G.
1 980. Transition and refractory element additions to rapidly solidified aluminum alloys.
Rapid Solidification Processing: Principles
and Technologies II, eds. R. Mehrabian, B.
H. Kear and M. Cohen, Claitors Publishing,
Baton Rouge, La., pp. 246-259.
4. Adam, C. M. 1982. Rapidly Solidified
Amorphous and Crystalline Alloys, Vol. 8,
eds. B. H. Kear, B. C. Giessen and M. Cohen,
Elsevier Science Publishing Co. New York,
pp. 411^122.
5. Anderson 1. E., and Kempainen, M. P.
1988. Undercooling effects in gas-atomized
powders. Undercooled Alloy Phases, eds. E.
W. Collings and C. C. Koch, TMS, Warrendale, Pa, pp. 269-285.
6. Adam, C. M. 1986. Rapid solidification
technology: The potential for innovative alloy
properties. Science and Technology of Undercooled Melts, eds. P. R. Sahm, H. Jones
and C. M. Adams, NATO-ASI Series, Martinus-Nijhoff Publishers, Dodrecht, pp.
1 86-209.
7. Adam, C. M. 1985. Rapid solidification
processing methods. Mechanical Behavior of
Rapidly Solidified Materials, eds. S. M. L. Sastry and B. A. MacDonald, TMS, Warrendale,
Pa., pp. 21-39.
8. Baeslack III, W. A. 1985. Microstructural transitions in a welded powder metallurgy AI-8Fe-1.7Ni alloy. Metallography, 18,
pp. 73-82.
9. Savage, S. J., and Froes, F. H. 1984.
Production of rapidly solidified metals and
alloys, journal of Metals, 36 (4): 20-33.
10. Trebbia, P. 1989. CLEDX: a standalone program for quantitative x-ray analysis
of thin films using the Cliff-Lorimer procedure. Ultramicroscopy, 27: 343-348.
11. Mondolfo, L. F. 1976. Aluminum Alloys: Structure and Properties. Butterworths,
London, England, pp. 283-289.
12. Mondolfo, L F. 1 976. Aluminum Alloys: Structure and Properties. Butterworths,
London, England, pp. 329-331.
13. Jones, H. 1982. Rapid Solidification
of Metals and Alloys. Monogr. 8, Institution
of Metallurgists, London, England, pp.40-43.
14. Armstrong, G. R., and Jones, H. 1979.
Effect of decreased section thickness on the
formation, structure and properties of a chill
cast Al-Si alloy. Proc. Conf. Solidification and
Casting of Metals, The Metals Society, London, England, pp. 454-459.
15. Jones, H. 1982. Rapid Solidification
of Metals and Alloys. Monogr. 8, Institution
of Metallurgists, London, England, pp. 45.
1 6. Kurz, W., and Fisher, D. J. 1986. Fundamentals of Solidification, Trans Tech Publications, Aedermannsdorf, Switzerland, pp.
67.
17. Perepezko, J. H., and LeBeau, S. E.
1 982. Nucleation of Al during solidification,
Aluminum Transformation Technology and
Applications —1981, eds. C. A. Pampillo, H.
Biloni, L. F. Mondolfo and F. Sacchi, ASM,
Materials Park, Ohio, pp. 309-346.
1 8. Jones, H. 1 969/70. Observations on
a structural transition in aluminum alloys
hardened by rapid solidification. Materials
Science & Engineering, pp. 1-18.
WELDING RESEARCH SUPPLEMENT I 427-s