Heat Treatments for Improving the Weldability and

Formability of Udimet 700

Fabricability of Udimet 700 sheet is improved by heat treatments

which overage the strengthening of y' precipitate, and maximum
resistance to cracking during welding and postweld heat treatment
is obtained by a two-step overaging treatment
BY D. S. D U V A L L A N D W. A. OWCZARSKI

ARSTRACT. This study was initiated to de- materials have frequently been de- cracking." Age-hardenable nickel-base
termine if the poor weldability and form- rived from w r o u g h t nickel-base super- alloys must be heat treated after weld-
ability of the nickel-base superalloy Udi- alloys used successfully in forgings. ing to restore mechanical properties
met 700 could be significantly improved O n e alloy whose properties show and relieve residual stresses. Alloys
by alteration of its microstructure and
properties through heat treatment. Speci- promise as a potential advanced sheet such as Waspaloy and R e n e 41 gener-
mens given several experimental heat material is U d i m e t 7 0 0 ( A s t r o l o g y ) . * ally exhibit "strain-age" cracking only
treatments designed to produce the de- Evaluations of limited quantities of when welded in the fully heat-treated
sired microstructural features were evalu- sheet material indicate that this alloy condition (either initially or during
ated by bend, hardness and tensile tests, has elevated t e m p e r a t u r e mechanical repair welding) and then taken direct-
and electron metallography. Promising properties superior to those of any ly to the aging t e m p e r a t u r e during the
heat treatments which reduced strength superalloy sheet in widespread use to- following heat treatment. 2 ' s Re-
and increased ductility were chosen for
day. While these properties m a k e the strained U d i m e t - 7 0 0 weldments, how-
use in subsequent welding experiments.
Weldability studies were conducted on alloy an attractive choice for ad- ever, generally cannot be postweld
sheet given the various heat treatments vanced sheet c o m p o n e n t s , its use in heat treated without cracking even
using a modified restrained-weld patch sheet form has been restricted by two when welded in the solutioned condi-
test. f a c t o r s — p o o r weldability and limited tion and then heated as rapidly as
The fabricability of Udimet-700 sheet formability. possible through the aging range up to
was substantially improved by heat treat- U d i m e t 7 0 0 is susceptible to hot the solution heat-treatment tempera-
ments which overaged the strengthening cracking in both the weld and heat- ture. 2
7' precipitate. Maximum resistance to affected zone during welding. 1 A n o t h - A n o t h e r undesirable characteristic
weld cracking both during welding and er weldability problem with this alloy of U d i m e t - 7 0 0 sheet is the alloy's
during postweld heat treatment was ob- is its tendency to crack during post- relatively poor formability. T h e capa-
tained by a two-step overaging treatment
weld heat t r e a t m e n t u n d e r the influ- bility of being easily formed into
(solution—2140° F / 4 hr; age 1975° F /
16 hr plus furnace cool to 1850° F, hold ence of residual stress, i.e., "strain-age h a r d w a r e shapes at low temperatures
4 hr then slow cool). The same two-step is an i m p o r t a n t characteristic for a
overaging treatment followed by oil useful sheet material. Yet, solution
quenching from 1850° F produced addi- Table 1—Chemical Analyses of Alloys heat-treated U d i m e t 700 has substan-
tional improvement in room temperature Used in This Study tially less capacity for deformation at
formability. Bend ductility was increased 5/8 in. r o o m t e m p e r a t u r e than other widely
nearly three-fold over that in the solution diameter 0.060 inch used superalloy sheet materials.
heat-treated alloy. The enhanced fabric- barstock, thick sheet, In order to improve the weldability
ability afforded by the overaging heat wt-% Element wt-%
treatments should permit application of and formability of U d i m e t - 7 0 0 sheet,
high-strength superalloys such as this al- Balance Nickel Balance several of its properties needed to be
loy to many components which previously 15.4 Chromium 14.4 modified. A microstructure was
could not be fabricated out of these ma- 5.0 Molybdenum 4.5 sought in the alloy which had low
terials. 18.8 Cobalt 16.7 yield strength and high ductility at
4.4 Aluminum 3.6
room and elevated t e m p e r a t u r e s . F u r -
3.4 Titanium 3.3
Introduction 0.06 Carbon 0.08 t h e r m o r e , in o r d e r to reduce the tend-
0.03 Boron 0.04 ency for postweld heat t r e a t m e n t
High-strength, heat-resistant sheet
0.13 Iron 0.05
D. S. DUVALL and W. A. OWCZARSKI 0.006 Sulfur Not deter- •Udimet 700 and Astrology are similar
are with Pratt & Whitney Aircraft, Mid- mined in composition (the chemistries of a heat
dletown. Conn. of Udimet 700 and a heat of Astrology are
Paper sponsored by Welding Research 0.10 max. Silicon 0.05 listed in Table 1). Throughout this paper
Council and presented at the AWS 52nd 0.10 max. Manganese 0.01 max. the term Udimet 700 will be used except
Annual Meeting held in San Francisco, 0.05 max. Zirconium 0.05 max. where significant differences are felt to be
Calif., during April 26-29, 1971. important.

WELDING RESEARCH SUPPLEMENT | 401-s

 Table 2—Room Temperature Bend Ductility and Hardness After Various
Heat Treatments
Bend Hardness
History angle, deg VHN
Solution heat treated 64 380
(2140° F/4 hr + forced air cool)
Fully Heat Treated 24 407
(2140° F/4 hr + 1975° F/4 hr + 1550° F/4 hr +
1440° F/16 hr)
Fig. 1—Restrained-weld strip patch-test One-step overaging treatments:
specimen (arrows indicate location of Solution heat treat plus following:
test weld) 1975° F/l hr plus oil quench 115
1975° F/4 hr plus oil quench 137
1975° F/16 hr plus oil quench 147 320
cracking, the microstructure produced 1975° F/64 hr plus oil quench 145
should not precipitate additional 1975° F/16 hr plus furnace cool 73
phases such as the strengthening pre- 1900° F/16 hr plus oil quench 80
cipitate y or deleterious grain- 1850° F/16 hr plus oil quench 47
boundary carbides during the heating Two-step overaging treatments:
Solution heat treat plus 1975° F/16 hr plus
stage of a postweld heat treatment.
furnace cool to following:
The investigation described in this pa- 1900° F/4 hr plus oil quench 142
per was conducted to seek heat treat- 1850° F/4 hr plus oil quench 168 290
ments which would produce this de- 1850° F/l hr plus oil quench 121
sired microstructure thereby enhanc- 1800° F/4 hr plus oil quench 154
ing the overall fabricability of Udi- 1700° F/4 hr plus oil quench 106
met 700 sheet. The work was con- Two-step overaging treatment plus slow cool:
ducted in two phases: Solution heat treat plus 1975° F/16 hr plus 74 344
furnace cool to 1850° F/4 hr plus slow cool to room
1. Selection of the optimum heat temperature (maximum 100° F/hr to 1050° F, furnace
treatments based upon a study of the cool to room temperature)
effects of various heat treatments on
alloy properties.
2. An evaluation of the degree of ure of bend ductility. Several of the made in the center of the strip after it
improvement imparted to weldability more ductile samples did not crack had been welded to the "strong back"
and formability by the optimized heat when bent to the maximum amount at both ends. The strip-patch test
treatments. (90 deg) permitted by the fixture. In configuration is illustrated in Fig. 1.
order to obtain a measure of the All welds were made by automatic gas
Materials and Experimental relative ductilities of these particular tungsten-arc welding with filler metal
Procedures samples, they were removed from the additions. The welding conditions are
Specimens used for tensile, bend, bend fixture and bending was contin- discussed below.
and hardness tests, and metallographic ued to failure by compressing the ends
studies were produced from 5 / 8 in. of the specimen together in a vise. Results
diameter, hot-rolled, centerless-ground Hardness measurements were made
barstock. Tensile specimens were tak- on metallographic samples using a Heat-Treatment Evaluation
en directly from the 5 / 8 in. diameter Vickers hardness tester with a 30 kg The initial evaluation of the experi-
bar, while the barstock was rolled (at load. Reported values are the average mental heat treatments consisted of
1925° F) into V 8 in. thick slabs to of five readings. Room and elevated- bend tests to determine the conditions
provide material for bend and hard- temperature tensile tests were con- which produced maximum room-
ness specimens and metallographic ducted on selected specimens at a temperature ductility. These heat
samples. The composition of this heat strain rate of 0.033 min"1. Micro- treatments were designed to develop a
of the alloy is listed in Table 1. All structural changes produced by the coarse, overaged dispersion of the
weldability studies were made on various heat treatments were observed strengthening •/ precipitate which,
0.060 in. thick, commercially-rolled by light and electron metallography. when retained at room temperature,
sheet. Its composition is also given in Sample preparation techniques have
would give optimum ductility and a
Table 1. been described elsewhere. 1
low yield strength. The room-
The effects of various heat treat- Weldability of the sheet in various temperature bend ductilities obtained
ments were studied by means of bend heat-treated conditions was investi- following various heat treatments are
tests, hardness tests, tensile tests, and gated by means of restrained-weld listed in Table 2. Shown for compari-
light and electron metallography. patch tests. The restraint was provided son are the bend ductilities of fully
Bend tests were conducted at room by initially welding the sheet test spec- heat-treated and solution heat-treated
temperature using a guided bend imen to a Hastelloy-X "strong back" material whose microstructures are
fixture with a 0.062 in. (0.5T) radius before making the test weld on the shown in Figs. 2 and 3.
plunger. Specimens were approx- specimen. The "strong backs" were The initial group of bend specimens
imately 1.5 in. long, 0.2 in. wide, and 6V S in. diameter rings, 1V 4 in. were solution heat treated (2140°
0.125 in. thick. Bend tests were termi- thick, with a 2 3 / 4 in. diameter center F / 4 hr + forced air cool) and then
nated at the onset of cracking when hole. A strip-patch weld specimen was given a single-step averaging treat-
possible, although some of the less chosen which consisted of a 3 to 3V 4 ment at temperatures from 1850 to
ductile specimens failed completely in. long by 1 in. wide strip of 0.060 1975° F for times of 1 to 64 hr. Bend
before the test could be terminated. inch thick sheet. Longitudinal (2V 4 ductilities of these specimens varied
The bend angle achieved when crack- in. long) or transverse (of varying from 47 to 147 deg with the max-
ing commenced was taken as the meas- length) bead-on-plate test welds were imum ductility produced by the 1975°

402-s SEPTEMBER 1971

 • •

. ' . ' • • •

:
. - • • :•.••••• ' - K 7 • " • " • • • * : . . • - •• •.

..
• > ' " . • -
.. • ••.. •••
-
••
". ' - -\ •
- ' . .
• • • - : .
,
v
•• •:
'
•••••.•
* •
< , .
• •
•
:
. . .
. .
•'
)
•-•: '777.,. •-":,*,' *"> '•

•-••< •••••.-:% - ' . • ' • • - • . . . . • • < • - • , •••'• •••

- • ' • • • . ^

•i 1
Fig. 2—Electron micrograph of solution Fig. 3—Electron micrograph of fully Fig. 4—Electron micrograph of material
heat-treated material. X10.000 heat-treated material. X10.000 given single step overage (solution heat
treatment + 1975° F/16 hr/oil quench).
X10,000
F / 1 6 hr/oil quench treatment. The
electron micrograph in Fig. 4 shows cooling (arbitrarily selected at 50°
that the microstructure produced by F / h r to 1650° F; then 100° F / h r to
this thermal treatment consists of 1050° F; then furnace cooled to room
coarse y particles dispersed through- temperature) induced further precipi-
out the grains and along grain bound- tation of y and reduced the ability of
aries. No MogCo-type grain-boundary such a specimen to precipitate addi-
carbides were visible in this structure. tional y during subsequent heating.
Compare this to the duplex (coarse + As anticipated, however, the slow
fine) y structure observed in the fully cooling lowered the room-temperature
heat-treated alloy (Fig. 3) and the bend ductility due to the precipitation
fine dispersion of y found in normal boundary carbides. The samples which
solution heat-treated Udimet 700 were slow cooled from 1850° F were
(Fig. 2). somewhat more ductile at room tem-
The next series of heat treatments J. perature (bend angle + 74 deg) than
were designed to cause further precip- either solution heat-treated (64 deg)
Fig. 5—Electron micrograph of material
itation of y on the existing particles or fully heat-treated material (24
given two-step overage + rapid cool
established during the 1975° F / 1 6 hr (solution heat treatment + 1975° F/16
deg), but they were considerably less
exposure to reduce the matrix hr + cool to 1850° F, hold for 4 hr + ductile than those oil quenched from
strengthening and promote additional oil quench). X10,000 1850° F (168 deg). The hardness of
thermal stability of the precipitate the slow-cooled specimens (344
structure. Additional samples were so- f VHN) was less than that of the solu-
lution heat treated, aged at 1975° / tion heat-treated (380 VHN) or fully
F / 1 6 hr, and then furnace cooled (at heat-treated material (407 V H N ) ,
~ 100° F/hr) directly to a second but higher than the oil quenched sam-
aging temperature (which was varied ples (290 V H N ) . The microstructure
from 1700 to 1900° F) and held for produced by slow cooling following
different times prior to oil quenching. the two-step treatment is shown in
Bend-test results for these specimens Fig. 6. The fine y' precipitate in the
are also listed in Table 2. The max- matrix and some grain boundary
imum room-temperature bend ductili- M23C(i-type carbides can be seen.
ty obtained in the entire study was Room-temperature tensile tests
produced by one of these two-step •• were conducted on several specimens
heat treatments—furnace cooling in order to more quantitatively assess
from 1975° F (after 16 hr at temper- the effects of the most promising over-
ature) to 1850° F, holding for 4 hr • .' aging thermal treatments on mechani-
and then oil quenching. The structure J cal properties. These data are
developed by this treatment is shown presented in Fig. 7. The room-
Fig. 6—Electron micrograph of material
in Fig. 5. As expected, the volume given two-step overage -f slow cool temperature tensile properties of the
fraction of y is increased compared (solution heat treatment + 1975° F/16 solution and fully heat-treated condi-
to that following the single-step treat- hr + cool to 1850° F, hold for 4 hr + tions are compared with the proper-
ment (shown in Fig. 4 ) . Again, no very slow cool). X10.000
M ties of the alloys in the two conditions
23 c (rtype grain-boundary carbides
are visible in this microstructure illus- of improved fabricability, i.e., solution
trated in Fig. 5. tion during the initial part of a post- heat treatment plus two-step overage
weld heat treatment. In order to re- plus either oil quench or very slow
Although the two-step overaging duce this subsequent precipitation po- cool from 1850° F. In addition, ten-
treatment plus rapid cool described tential (and improve postweld heat- sile samples given the two-step over-
above developed the greatest room- treatment cracking resistance), sam- age plus oil quench were then fully
temperature ductility, the microstruc- ples were very slowly cooled instead heat treated (i.e., re-solution plus
ture created still had the potential for of oil quenched from 1850° F follow- multi-step aging) to assure that
additional y and carbide precipita- ing the two-step overaging. The slow mechanical properties could be

WELDING RESEARCH SUPPLEMENT | 403-s

 139.0
DUCTILITY

30 28.2
5
I-
O
z
O 20

1
18.6

STRENGTH

Fig. 7—Room temperature ten-

205 202 sile properties of barstock
given various heat treatments
189
181

ULTIMATE

140

125

100

SOLUTION SOLN.+ SOLN.+ NORMAL SOLN. + 2-STEP

HEAT 2-STEP 2-STEP FULL OVERAGE-OIL QUENCH
TREATMENT OVERAGE+ OVERAGE + HEAT + FULL HEAT
OIL QUENCH SLOW COOL TREATMENT TREATMENT

restored in material subjected to fabri- resistance to postweld heat-treatment testing.

cation-optimizing heat treatments. cracking. Both heat treatments were The weld tests evaluated the effects
It is apparent from Fig. 7 that both selected for further studies to deter- of the four selected heat treatments
thermal treatments for promoting mine their effect on the weldability of (i.e., solution, full, two-step overage
ductility produce more room- this alloy. To provide comparative plus rapid cool and two-step overage
temperature elongation (39.0% for data, the solution and fully heat- plus slow cool) on the hot-cracking
oil-quenched material, 33.0% for treated conditions were also included and post-weld heat-treatment cracking
slow-cooled specimens) than that ex- in the weldability investigation. The sensitivity of the sheet. The largest
hibited by either fully heat-treated heat treatments selected for welding group of test welds were made using
(19.4 % ) or solution heat-treated ma- studies are listed in Table 3. Waspaloy filler metal* with the weld-
terial ( 2 8 . 2 % ) . Furthermore, the ing parameters adjusted to produce
yield strengths are reduced substan- Weldability Evaluation approximately 7 5 % weld penetration.
tially by these treatments thus making The weldability tests were conduct- Following postweld visual and dye-
the alloy easier to form. It can also be ed with the strip-patch test specimen penetrant inspection, the strong backs
seen from these data that tensile prop- shown in Fig. 1. This specimen was were thermocoupled, and most speci-
erties are restored to normal values constructed by first welding the ends mens were then heated at ~ 3000°
when samples are fully heat treated of the strip to the strong back and F / h r to 1975° F in an argon atmos-
subsequent to the overaging heat subsequently making the test weld in phere. The samples were normally
treatments. the center of the specimen. Test welds removed from the furnace immediate-
were initially made in the transverse ly upon reaching this temperature and
The results of this heat-treatment (i.e., across the width of the strip) examined for signs of cracking.
evaluation suggested that the two-step direction. Further tests showed, how-
overaging thermal treatment followed The weldment specimens were
ever, that the desired degree of test
by a rapid cool would best condition qualitatively rated as to the degree of
severity could best be obtained by
Udimet 700 for maximum formabili- both hot cracking and postweld heat-
making the bead-on-plate test weld
ty. It was felt that the modified form longitudinally (i.e., parallel to the treatment ("strain-age") cracking
of this heat treatment (employing a length direction of the strip). Conse-
very slow cool after the two-step quently, this configuration was used •Composition (wt-%) as follows: Cr—
19.3. Co—13.5. Ti—3.0, Al—1.4, Mo—4.0,
overaging) might impart additional for the balance of the weldability C—0.1, Ni—balance.

404-s SEPTEMBER 1971

 RELATIVE DEGREE OF RELATIVE
STRAIN-AC E CRACKING CRACK SENSITIVITY
HEAT-
TREATED STRAW-
CONDITION DURINC ACE
NONE SLIGHT MODERATE 5EVERE WELDING CRACKING

SOLUTION
MEAT
• • HIGH
HIGH
TREATMENT

FULL
HEAT
• r A INTER. INTER-
TREATMENT MEDIATE MEDIATE

SOLUTION
+ 2-STEP
O.VERACE
• LOW HIGH
+ OIL QUENCH

SOLUTION
+ 2-STEP
• • mA • s LOW
OVERAGE LOW
H-SLOW COOL
A
| INITIAL WELD (WASPALOY FILLER)

^ SIMULATED REPAIR WELD (WASPALOY FILLER)

f/k INITIAL WELD (INCONEL 71S FILLER)

A 10WF/MOUR HEATING RATE

Fig. 9—Cracking in strip patch-test spec-
8 FULL PENETRATION WELD imens following postweld heat treat-
Fig. 8—Weldability test results for various heat-treated conditions of ment. A (top)—solution heat treated
0.060 in. thick sheet material (Waspaloy filler metal), SEVERE crack-
ing; B (bottom)—two-step overaged -f
which were encountered. The relative slow cooled (Waspaloy filler metal),
ence in the degree of cracking be- SLIGHT cracking (arrow)
crack susceptibilities of material given tween the two conditions can be seen.
the various preweld heat treatments In order to differentiate between
are summarized in Fig. 8. It was the strain-age crack susceptibility of 718, it was felt that dilution of the
observed that specimens given either the fully heat-treated and overaged, weld metal with this composition
of the two-step overaging treatments slow-cooled conditions, two specimens might reduce the cracking tendencies
(i.e., rapid cool and slow cool) had were heated at a slower rate of ~ of the Udimet 700 weldments. As
the least tendency to crack during 1000° F / h r to 1975° F. The greater shown in Fig. 8, the overaged plus
welding. The fully heat-treated sam- severity of the test caused by the slow-cooled specimen welded with Al-
ples had a somewhat greater tendency reduction in heating rate (hence a loy 718 filler metal was successfully
towards hot cracking, while the solu- longer time in transit through the postweld heat treated (at 3000°
tion heat treatment produced the heat-treatment cracking temperature F/hour) without cracking. No im-
greatest hot-crack sensitivity in the range) increased the amount of provement was noted, however, in the
alloy. When hot cracking was encoun- cracking in the fully heat-treated spec- solution heat-treated specimen welded
tered, it principally occurred in the imen to the moderate level. However, with this filler metal, and severe
heat-affected zone although some the overaged plus slow-cooled spec- cracking still occurred.
cracks extended into the weld metal. imen still exhibited only slight crack-
However, in all cases the degree of Discussion
ing. Simulated repair welding also
hot cracking was less severe than the demonstrated the difference in "strain- This investigation has shown that
weldment cracking which took place age" crack sensitivity between the ful- both weldability and formability of
during subsequent postweld heat treat- ly heat-treated alloy and the material Udimet 700 sheet can be improved by
ment. given the two-step-overage plus slow- means of preweld heat treatments.
The relative amounts of postweld cool treatment. A welded test speci- The room-temperature bend ductility
heat-treatment cracking which were men in each of the two conditions was of this material is greatly increased by
observed are also compared in Fig. 8. subjected to a simulated "repair" by the two-step overaging heat treatment
The degree of cracking was qualita- making a transverse weld near one followed by a rapid cool. In addition,
tively rated on an arbitrary scale of end of the original, longitudinal test the postweld heat-treatment crack
None, Slight, Moderate or Severe. weld. These specimens, each contain- sensitivity is reduced through the use
Under these test conditions, both the ing a "T"-shaped weld, were then of these heat treatments combined
solution heat-treated and the two-step heated at 3000° F / h r to 1975° F. with a slow cool. The resistance to
overage plus rapid-cool specimens While the overaged plus slow-cooled heat-affected zone cracking during
cracked severely. The fully heat- specimen again only cracked slightly, welding is enhanced by the two-step
treated specimen and that given the the fully heat-treated alloy was overaging treatment with either a fast
two-step overage plus slow cool, how- severely cracked. or slow cool. These improvements to
ever, were only slightly cracked. A The effect of filler metal composi- fabricability have been achieved by
solution heat-treated specimen (severe tion was briefly examined by testing alterations of the microstructure
cracking) and one given the two-step two specimens welded with Alloy 718
overage plus slow cool (slight crack- filler metal.f Because of the low t C o m p o s i t i o n (wt-%) as follows: Cr—
18.5. Fe—18, Cb—5, Mo—3, Ti—0.9, Al—
ing) are shown in Fig. 9. The differ- "strain-age" crack sensitivity of Alloy 0.6, C—0.1, Ni—balance.

WELDING RESEARCH SUPPLEMENT | 405-s

 HU I l I

60 - -
PERCENT
OVERAGE PLUS SLOW COOL
REDUCTION -
40
OF
iy "
AREA S ^FULL H.T. \ \
20
SOLUTION H . T \

o i I i~~--A.
1400 1600 1800 2000 22
TEMPERATURE ~ ° F
Fig. 10—Synthetically p r o d u c e d heat-affected z o n e h o t - d u c t i l i t y data.
Tests c o n d u c t e d on h e a t i n g t o peak t e m p e r a t u r e s i n d i c a t e d . Data f o r
f u l l and s o l u t i o n h e a t - t r e a t e d m a t e r i a l taken f r o m O w c z a r s k i et a/ 1

which increase ductility and lower fully heated alloy (Table 2 and Fig. 1 Sli '
strength. 7) and, therefore, better room- >_ > •

Because of the material's high alloy temperature formability. However,

content, this improved microstructure synthetically produced heat-affected-
is better obtained by overaging the zone hot-ductility data (Fig. 10) have
strengthening y precipitate rather shown that the solution heat-treated
than attempting to retain the solute in material has considerably less tensile
solution (as is the case for less highly ductility than the fully heat treated
:.' ' - " ' ' " v'..;:••'/•"•• ,< V ; N V * > X . t i-l"
alloyed materials. With most precipi- alloy at elevated temperatures up to
tation hardened materials and with the y-y' solvus. This poor high- F i g . l l — Surfaces of w e l d s p e c i m e n s
s h o w i n g plastic d e f o r m a t i o n w h i c h oc-
the leaner nickel-base superalloys, temperature behavior contributes to
c u r r e d in heat-affected zones d u r i n g
maximum fabricability is achieved in the high sensitivity for heat-affected- w e l d i n g . S p e c i m e n s u r f a c e s were m e t a l -
the solution heat-treated condition. In zone hot cracking during welding. l o g r a p h i c a l l y p o l i s h e d t h r o u g h 2/J. dia-
this type of higher strength superalloy, The low ductility at these tempera- m o n d paste prior to w e l d i n g in o r d e r to
it is almost impossible to suppress y' tures is apparently caused by the ab- reveal s l i p traces. A ( t o p ) — t w o - s t e p
precipitation on cooling after solution sence of large grain-boundary parti- overaged + oil q u e n c h ; B ( b o t t o m ) —
heat treatment (even by rapid quench- cles which are needed to inhibit grain- fully heat t r e a t e d
ing) because of the high degree of boundary sliding.4 The y'-strength-
alloy supersaturation coupled with the ened grains resist deformation, while solution heat-treated alloy in the
inherently fast y formation kinetics. the precipitate-free grain boundaries 1400° F range is much less than that
Therefore, instead of being soft and can slide relatively easily resulting in of the fully heat-treated alloy. This
ductile, the "solution" heat-treated mi- stress concentrations and intergranu- correlates with the higher sensitivity
crostructure consists of relatively lar crack initiation in low macroscopic for postweld heat-treatment cracking
high-strength grain matrices contain- plastic strains. In the fully heat- exhibited by material in the solution
ing a large volume fraction of y treated alloy, the large intergranular heat-treated condition (Fig. 8).
precipitate in combination with rela- y particles are effective obstacles to Since it is impossible to satisfactori-
tively weak, precipitate-free grain grain-boundary sliding; deformation is ly prevent on-cooling y precipitation
boundaries (Fig. 2 ) . therefore more uniformly distributed in the solution heat-treated alloy, the
Solution heat-treated material has throughout the grains and high- best way to soften the material for
somewhat lower strength and greater temperature ductility is greater. Data 5 easier fabrication is to coarsen the
ductility at room temperature than the also show that the ductility of the existing y particles by overaging.
Overaging increases fabricability in
two ways. First, it lowers the strength
Table 3—Principal Heat Treatments Used for Weldability Studies of the alloy. The material can there-
fore be more easily formed and, dur-
Solution heat treat Full heat treatment
ing welding, any stresses which are
2140° F/4 hr + f o r c e d air cool 2140° F/4 hr + f o r c e d air cool generated can be more readily accom-
1975° F/4 hr + forced air cool modated by plastic yielding of the
1550° F/4 hr + f o r c e d air cool metal.
1400° F/16 hr + forced air cool
Two-step overage for Two-step overage for
Figure 11 illustrates the surface slip
maximum formability maximum weldability deformation observed in the heat-
2140° F/4 hr + f o r c e d a i r cool 2140° F/4 hr + f o r c e d a i r cool affected zones of autogenous welds
1975° F/16 hr -f- f u r n a c e cool at 1975° F/16 hr + f u r n a c e cool at made on mechanically polished sheet
100° F/hr t o : 100° F/hr t o : samples of fully heat-treated and on
1850° F/4 hr + oil q u e n c h 1850° F/4 hr + slow cool at 50°F/hr overaged Udimet 700. The plastic de-
f r o m 1850 to 1650° F; at 100° F/ hr formation induced by welding is easily
f r o m 1650 to 1050° F; f u r n a c e cool
revealed on the polished surfaces, and
f r o m 1050° F to r o o m t e m p e r a t u r e
it can be seen by the higher slip-line

40B-S I S E P T E M B E R 1971
 8 16 24 32 40 48 56 64 72
TIME-HRS
Fig. 12—Calculated changes in strength as a function of aging time for
single-step overaging at the indicated temperatures. Also shown are the Fig. 13—Electron micrograph of thin
equilibrium volume percentages of y present at each temperature foil from material oil quenched after
aging at 1975° F. Note copious amount
of fine "on-cooling" y interspersed
density that greater deformation has overaging temperatures involved) to among the large y particles. X50.000
concentrated in the heat-affected zone permit dislocations to bypass the pre-
of the fully heat-treated specimen cipitates by bowing, a modified
(Fig. 11B) compared to the overaged Orowan expression: 6 rentheses are the volume percentages
sample (Fig. 11 A ) . A second advan- of y present at each of the overaging
tage of overhanging is that it improves K temperatures. These curves confirm
ductility by weakening the precipita- Ar In r
R what was intuitively known, that the
tion-strengthened grain matrices while lowest strengths are obtainable at the
retaining the large intergranular y relates strength (^ T ) to the mean highest overhanging temperatures (i.e.,
particles which beneficially alter high- planar interparticle spacing (2R) and 1975° F) because of the lower vol-
temperature grain-boundary sliding. the mean planar particle radius (r) ume fraction and more rapid coarsen-
The results which we have discussed where K is a constant. 2R can also be ing kinetics of the / .
demonstrate the advantageous effects related to r and the volume fraction
Figure 12 shows that the greatest
of such overaging heat treatments on (/) of precipitate: 7
reduction in strength occurs by the
fabricability. However, these data in- first 10-20 hr of overaging so that
dicate that the degree of improvement Kr little additional benefit is gained by
2R 2fU2
to weldability is very sensitive to the longer heat treating. (Note the simi-
exact time-temperature sequence of larity in bend ductility in Table 2
the overaging heat treatment which is In turn, between samples overaged 16 and 64
employed. The reasons why the condi- hr at 1975° F.) Also, the two-step
tioin of lowest strength is obtained K overage actually produces a more
with a two-step rather than a single- AT In r effective strengthening dispersion of
step overaging heat treatment is not at
r(f~ 1)
the large y particles which are pre-
first apparent. The reduction in Using this expression and the parti- cipitated at the overaging temperature
strength produced by a single-step cle size and volume fraction data compared to a 1975° F single-step
treatment can be approximately cal- measured for this alloy by Van Der treatment. (The calculated relative
culated as a function of the aging time Molen, et al.,8 the relative changes in strength following the two-step treat-
by considering the concurrent changes strength of this alloy with overaging ment is shown in Fig. 12.)
to the y particle size and interparticle at three different temperatures have While Fig. 12 indicates that the
spacing. Assuming that the volume been approximately calculated and are maximum softening can be obtained
fraction of y is low enough (at the shown in Fig. 12. Also shown in pa- by a single-step heat treatment at a

.,

..o.
^WM:M
•'•. %

100L
M' '•' '

Fig. 14—The as-welded heat-affected zone microstructures of material given various preweld heat treatments. A (left)—welded
in fully heat-treated condition; B (right)—welded following the two-step overaging heat treatment. X200 (reduced 26<% on re-
production)

WELDING RESEARCH SUPPLEMENT | 407-s

temperature approaching the alloy's affected zone microstructures of cooled. Replication electron microsco-
y-y solvus ( ~ 2050-2075° F ) , the weldments made in the fully heat- py showed that the two-step overage
experimental data demonstrate that a treated and overaged conditions. The plus slow cooled samples (Fig. 5 and
two-step overage is really more effec- similar appearance of the two heat- 6) contained grain-boundary carbides
tive. The room-temperature 0.2% affected zones, including the amount (presumably M 23 C |; ) which precipi-
yield strength of material overaged at of partial melting, can be seen. Note tated during the slow cool from 1850°
1975° F/16 hr + 1850° F / 4 hr/oil the absence of intergranular hot F. Specimens oil quenched from
quench was 104 ksi (Fig. 7) com- cracking in the overaged specimen 1850° F exhibited very little inter-
pared to 120 ksi for the 1975° F / 1 6 however. granular carbide precipitation. The
hr/oil quench treatment alone. The Although the two-step overaging oil-quenched material, therefore, had
reason for this discrepancy between treatment followed by oil quenching greater potential for carbon/carbide
calculation and experiment is the pre- greatly improved formability and reactions during subsequent thermal
cipitation of additional y on-cooling resistance to hot cracking during exposure such as a postweld heat
following overaging. Examination of welding, it failed to reduce the sensi- treatment. However, no direct evi-
thin foils by transmission microscopy tivity to postweld heat-treatment dence could be found which linked
revealed that oil quenching after over- cracking. In contrast, the substitution carbide precipitation with the poor
aging at 1975° F failed to suppress of a slow cool in place of oil quench- 1400° F ductility and "strain-age"
on-cooling y precipitation (Fig. 13). ing following the overaging signifi- crack sensitivity of the oil-quenched
Because only ~ 18 volume-% / is cantly alleviated both types of material.
formed by aging at 1975° F, the weldment cracking. The exact reasons This study has shown that specific
potential for considerable additional for this difference in behavior are not preweld overaging heat treatments
precipitation remains in that the max- clear. One effect of the slow cool was can reduce both the hot cracking and
imum amount of y obtainable below to permit the equilibrium volume postweld heat-treatment cracking
~ 1600° F is 3 8 % . In contrast, fraction of y to form in the base problems associated with Udimet 700
after aging at 1850° F, 33 volume-% metal. Consequently, during postweld weldments. Although overaged mate-
y has already formed and the super- heat treatment, the y' in the base- rial is still too "strain-age" crack sus-
saturation on cooling is greatly re- metal was either stable or undergoing ceptible to be fabricated in highly
duced. It has been showing that the dissolution rather than precipitating restrained, complex weldments, the
presence of the ~ 75-100A diameter (with the attendant hardening and use of the overaging heat treatments
on-cooling y can make a significant possible aging contractions). Another will allow expanded use in many weld-
contribution to the low-temperature important effect of the slow cool was ed components which previously could
•strength of superalloys, and the higher to improve the intermediate- not be fabricated out of this material.
•trength and lower ductility of the temperature ductility of the material. It also should be much easier to per-
ingle-step overaged material reflects While the oil-quenched specimens had form room-temperature forming op-
his. By two-step overaging at 1975 somewhat greater ductility at room erations on sheet material due to the
ind 1850° F followed by oil quench- temperature than those that were slow improved ductility and lower yield
ing, it is possible to obtain a satis- cooled (Fig. 7 ) , they had considerably strength of the alloy following the
factory compromise between the less ductility at 1400° F. Table 4 lists two-step overaging heat treatment and
weakening possible at elevated tem- the average 1400° F tensile properties the rapid coop. Prefabrication overag-
perature and the reduction of on- of oil-quenched samples compared to ing heat treatments can also be used to
cooling y such that the condition of ones given the slow cool and to fully improve the weldability of other high-
lowest strength and greatest ductility heat-treated Udimet 700. The 1400° strength wrought and cast superalloys.
at room temperature is achieved. F ductilities of these three base-metal However, this study illustrates that it
In addition to lowering the base- conditions correlate with the relative is necessary to tailor overaging heat-
metal strength, overaging also im- tendencies for heat-affected zone treatment cycles to the specific metal-
proves hot-cracking resistance by in- "strain-age" cracking (Table 4 ) . A lurgical characteristics of the alloy in
creasing the on-heating heat-affected similar correlation has been observed question.
zone ductility (Fig. 10), On-cooling between base-metal 1400° F tensile
from a peak temperature of 2200° F, properties and postweld heat- Conclusions
the properties of overaged heat- treatment cracking in Rene 41. 1 0 1. The weldability and formability
affected zones were identical to fully Examination of thin foils by trans- of Udimet 700 sheet can be substan-
heat-treated specimens. This is to be mission electron microscopy failed to tially improved by the use of specific
expected as the overaged alloy under- reveal any differences in the deforma- heat treatments which selectively alter
goes the same partial melting reac- tion behavior between the oil- the microstructural characteristics of
tions and dissolution of y in the que'n'ched and the slow-cooled tensile the alloy. These heat treatments over-
high-temperature heat-affected zone specimens. The duplex (large and age or coalesce the strengthening y
regions near the fusion zone as the small) y structure was similar in each precipitates to produce a microstruc-
material in any other heat-treated except that the fine y particles were ture that is softer and more ductile
condition. Figure 14 shows the heat- larger in the sample which was slow than the solution heat-treated condi-
tion previously employed.
2. The heat treatment producing
Table 4—Average 1400(' F Tensile Properties After Var ious Heat Treatments
the best overall weldability consists of
Relative " s t r a i n - a solution heat treatment (2140° F / 4
Heat-treated 0.2% yield Ultimate Elongation, a g e " cracking hr/forced air cool) followed by a
condition s t r e n g t h , ksi s t r e n g t h , ksi % sensitivity two-step overaging treatment (1975°
2-step overage 92 111 7 High F / 1 6 hr—cool at 100° F / h r to—
plus oil q u e n c h 1850° F / 4 hr) and a subsequent slow
2-step overage 86 126 32 Low cool (e.g., 50° F / h r from 1850 to
plus slow cool 1650° F : 100° F / h r from 1650 to
Full heat t r e a t m e n t 118 151 24 Intermediate
1050° F; cool to room temperature).

408-s SEPTEMBER 1971

 3. The heat treatment giving max- Acknowledgements Suppl., 17-s to 22-s (1962).
4. Oblak. J. M., Owczarski, W. A., a n d
imum room-temperature formability Duvall, D. S., " T h e R e l a t i o n s h i p of Micro-
The authors wish to thank G. C.
is the same two-step overaging treat- s t r u c t u r e to H o t W o r k a b i l i t y in a H i g h
Sikorowicz for his preparation and aid S t r e n g t h Nickel-Base S u p e r a l l o y , " in press,
ment for optimum weldability (see 2
in testing of the numerous weldability Met. Trans., May 1971.
above) except that a rapid cool (e.g., 5. Duvall. D. S., unpublished data.
specimens and E. T. Beebe for his
oil quench) is employed following the 6. Singhal, L. K.. a n d M a r t i n , J. W..
assistance with the experimental heat " T h e Mechanism of Tensile Yield in an
exposure at 1850° F. This heat treat-
treating. The electron metallography A g e - H a r d e n e d Steel Containing y (Ordered
ment increases the room-temperature NiaTi) P r e c i p i t a t e s , " Acta Met., 16(7), 947-
was conducted by J. M. Antol. 953 (1968).
bend ductility nearly three-fold over
7. Kelly, A., and Nicholson, R. B., " P r e -
that obtainable in the solution heat- cipitation H a r d e n i n g , " Prog. Mat'ls. Sci.,
treated condition. References
10(3). 149-391 (1963).
1. Owczarski, W. A.. Duvall, D. S.. a n d 8. Van Der Molen. E. H.. Oblak, J. M..
Sullivan, C. P., "A Model tor Heat-Affect- and Kriege, O. H.. "Control of y> P a r t i c l e
ed-Zone C r a c k i n g in Nickel-Base Super- Size and Volume F r a c t i o n in the H i g h
4. While this alloy is still suscepti- a l l o y s , " WELDING JOURNAL, 45(4), Research T e m p e r a t u r e Superalloy U d i m e t 700," in
Suppl.. 145-s to 155-s (1966). press, Met. Trans., 1971.
ble to postweld heat-treatment crack- 2. Lepkowski. W. J., Monroe, R. E., 9. B e a r d m o r e . P . , Davies, R. G., a n d
ing, the reduced weld crack sensitivity and Rieppel, P. J.. " S t u d i e s of R e p a i r J o h n s t o n , T. L... " O n e t h e T e m p e r a t u r e
and enhanced formability afforded by Welding A g e - H a r d e n a b l e Nickel-Base Al- Dependence of the Flow Stress of Nickel-
l o y s , " Ibid., 39(9), Research Suppl.. 392-s Base Alloys," Trans. AIME, 245(7), 1537-
the overaging heat treatments should to 400-s (1960). 1545 (1969).
permit its application in many com- 3. Weiss. S.. H u g h e s . W. P . . and Macke, 10. F a w l e y , R. W.. and P r a g e r . M.,
H. J., " W e l d i n g E v a l u a t i o n of H i g h T e m - " E v a l u a t i n g t h e Resistance of R e n e ' 41 to
ponents which previously could not be p e r a t u r e Sheet M a t e r i a l s by R e s t r a i n e d S t r a i n Age C r a c k i n g , " Welding Research
successfully welded. P a t c h T e s t i n g . " Ibid., 41(1), Research Council Bulletin No. 150, 1-12 (May 1970).

"Plastic Tests of Two Branch-Pipe Connections"

by N. C. Lind, A. N. Sherbourne, F. Ellyin, and J. Dainora

This investigation was undertaken to study the actual behavior of branch-pipe

connections in the inelastic range with a view to the ultimate development of a limit
analysis for these configurations.

The first configuration, designated as Model UW-J, was a standard reducing tee.
The second configuration, a standard full branch tee, was designated as UW-4. Both
vessels were Schedule 40, ASA B16.9, donated by Taylor Forge & Pipe Works.

"Bending of Pipe Bends with Elliptic Cross Sections"

by G. E. Findlay and J. Spence

A theoretical solution is presented for the flexibility and stresses in pipe bends
having elliptic cross sections under both in-plane and out-of-plane bending, when applied
separately. The theory is developed using an energy method and an assumed series
expression for the radial displacement, the assumptions being consistent with those in
the well-known Karman analysis for pipe bends with circular cross sections. Convergence
is obtained for both the flexibility and stress for all practical pipe bend parameters.

The relevance of the present theory to the bending of pipe bends with initial ovality
caused by manufacturing processes is discussed and simple design applications are
suggested.

Publication of the above papers was sponsored by the Pressure Vessel Research
Committee of the Welding Research Council. WRC Bulletin 164 is $2.00 per copy.
Orders should be sent to the Welding Research Council, 345 E. 47th St., New York,
N.Y. 10017.

WELDING RESEARCH SUPPLEMENT I 409-s