Solidification of Austenitic Stainless Steel Weldments: Part 2—The Effect of Alloy Composition on Ferrite Morphology A transition in ferrite morphology occurs as a function of composition and weld cooling rate BY J. C LIPPOLD AND W. F. SAVAGE ABSTRACT. The distribution and mor- phology of delta ferrite in duplex aus- tenitic stainless steel weldments has been shown to be a function of both the solidification process as in- fluenced by the composition of the alloy and the nature of the solid-state, ferrite-to-austenite transformation. Four separate regions have been iden- tified relative to a generalized pseudo- binary diagram of the Fe-Cr-Ni ternary system and used to explain the various duplex morphologies observed in the fusion zone. As the solidification mode shifts from one of primary austenite to one of primary ferrite (i.e., as the ratio of ferritizers to austenitizers increases), the location of the retained ferrite shifts from regions at the subgrain boundaries to the subgrain cores. A further increase in the Cr/Ni ratio results in a transition to an acicular austenite-ferrite morphology and fi- nally to a near-equilibrium mixture of Widmanstatten austenite along specif- ic habit planes in the ferrite. In addition to the alloy composition, the weld cooling rate was found to be an important factor influencing the ferrite morphology. Introduction The complex morphology of the duplex austenite-ferrite microstructure in austenitic stainless steel weldments has created considerable confusion as to the origin and nature of the residual delta ferrite. Although the composi- tions of both austenitic stainless steel base and filler metals are carefully balanced in order to produce a desired level of ferrite in the as-welded microstructure, the amount and distri- bution of ferrite can rarely.be pre- dicted accurately and often, both ex- hibit considerable variation within a weldment. In addition, the conditions of solidification of the alloy as influenced by both the choice of pro- cess and the welding parameters can significantly affect both the morphol- ogy and the amount of the retained ferrite. In an earlier report 1 the authors con- cluded, as had previous investiga- tors, 2 - 3 that: 1. Solidification of austenitic stain- less steel occurs as either primary delta ferrite or primary austenite. 2. Until the terminal transient peri- od, the solidification product was solely dependent on the ratio of ferrite stabilizers to austenite stabilizers in the melt. It was shown that, during solidifica- tion as primary austenite, chromium segregates to the liquid in advance of the solid-liquid interface. This contin- Paper presented at the AWS 60th Annual Meeting held in Detroit, Michigan, during April 2-6, 1979. /. C LIPPOLD, former Graduate Fellow at Rensselaer Polytechnic Institute, is with Sandia Laboratories, Livermore, California, and W. F. SAVAGE is Professor of Metallur- gical Engineering and Director of Welding Research, Rensselaer Polytechnic Institute, Troy, New York. ues until during the terminal transient period ferrite is formed as a divorced eutectic ferrite along subgrain bound- aries. Upon cooling from the solidifi- cation range, most of this ferrite remains stable and appears as a semi- continuous constituent along grain and subgrain boundaries in the as- welded microstructure. If the alloy is sufficiently rich in ferrite stabilizers, solidification occurs as delta ferrite and the first solid to form at the subgrain core is highly enriched in chromium and depleted in nickel. However, since welding is a classical case of nonequilibrium solid- ification and approximates Case III so- lidification conditions,* the bulk of the delta ferrite subgrain solidifies at the nominal composition of the alloy. Then, during the final stages of solidi- fication, a local enrichment of nickel results in the formation of austenite as a divorced eutectic constituent at the subgrain boundaries. Upon cooling to room temperature, the delta ferrite of nominal composi- tion is transformed to austenite by means of a massive transformation. The initial ferrite remains stable at room temperature and often appears in a semicontinuous network at the subgrain cores. An increase in the ratio of ferritizers to austenitizers results in a more extensive and continuous fer- rite morphology. *ln Case III solidification there is no diffu- sion in the solid and no mechanical mixing in the liquid. Concentration changes in the liquid occur by diffusion only. 48-sl FEBRUARY 1980
11
Embed
Solidification of Austenitic Stainless Steel Weldments ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Solidification of Austenitic Stainless Steel Weldments: Part 2—The Effect of Alloy
Composit ion on Ferrite Morphology
A transition in ferrite morphology occurs as a function of composition and weld cooling rate
BY J. C LIPPOLD AND W. F. SAVAGE
ABSTRACT. The distribution and morphology of delta ferrite in duplex austenitic stainless steel weldments has been shown to be a funct ion of both the solidification process as influenced by the composit ion of the alloy and the nature of the solid-state, ferr i te-to-austenite transformation. Four separate regions have been identified relative to a generalized pseudo-binary diagram of the Fe-Cr-Ni ternary system and used to explain the various duplex morphologies observed in the fusion zone.
As the solidification mode shifts from one of primary austenite to one of primary ferrite (i.e., as the ratio of ferritizers to austenitizers increases), the location of the retained ferrite shifts from regions at the subgrain boundaries to the subgrain cores. A further increase in the Cr/Ni ratio results in a transition to an acicular austenite-ferrite morphology and f i nally to a near-equilibrium mixture of Widmanstatten austenite along specific habit planes in the ferrite.
In addition to the alloy composit ion, the weld cooling rate was found to be an important factor influencing the ferrite morphology.
Introduction
The complex morphology of the duplex austenite-ferrite microstructure in austenitic stainless steel weldments has created considerable confusion as to the origin and nature of the residual delta ferrite. Although the compositions of both austenitic stainless steel
base and filler metals are carefully balanced in order to produce a desired level of ferrite in the as-welded microstructure, the amount and distribution of ferrite can rarely.be predicted accurately and often, both exhibit considerable variation wi th in a weldment. In addit ion, the conditions of solidification of the alloy as influenced by both the choice of process and the welding parameters can significantly affect both the morphology and the amount of the retained ferrite.
In an earlier report1 the authors concluded, as had previous investigators,2-3 that:
1. Solidification of austenitic stainless steel occurs as either primary delta ferrite or primary austenite.
2. Unti l the terminal transient period, the solidification product was solely dependent on the ratio of ferrite stabilizers to austenite stabilizers in the melt.
It was shown that, during solidification as primary austenite, chromium segregates to the liquid in advance of the solid-l iquid interface. This cont in-
Paper presented at the AWS 60th Annual Meeting held in Detroit, Michigan, during April 2-6, 1979.
/. C LIPPOLD, former Graduate Fellow at Rensselaer Polytechnic Institute, is with Sandia Laboratories, Livermore, California, and W. F. SAVAGE is Professor of Metallurgical Engineering and Director of Welding Research, Rensselaer Polytechnic Institute, Troy, New York.
ues until during the terminal transient period ferrite is formed as a divorced eutectic ferrite along subgrain boundaries. Upon cooling from the solidif ication range, most of this ferrite remains stable and appears as a semi-continuous constituent along grain and subgrain boundaries in the as-welded microstructure.
If the alloy is sufficiently rich in ferrite stabilizers, solidification occurs as delta ferrite and the first solid to form at the subgrain core is highly enriched in chromium and depleted in nickel. However, since welding is a classical case of nonequil ibrium solidification and approximates Case III solidification conditions,* the bulk of the delta ferrite subgrain solidifies at the nominal composition of the alloy. Then, during the final stages of solidification, a local enrichment of nickel results in the formation of austenite as a divorced eutectic constituent at the subgrain boundaries.
Upon cooling to room temperature, the delta ferrite of nominal composition is transformed to austenite by means of a massive transformation. The initial ferrite remains stable at room temperature and often appears in a semicontinuous network at the subgrain cores. An increase in the ratio of ferritizers to austenitizers results in a more extensive and continuous ferrite morphology.
*ln Case III solidification there is no diffusion in the solid and no mechanical mixing in the liquid. Concentration changes in the liquid occur by diffusion only.
48-s l FEBRUARY 1980
Table 1—Welding Parameters
1. 310/304 or 312/304L pulsed - GMAW pass.
2. Weld reinforcement removed.
GTAW pass to control ^-nal d i lu t ion.
Fig. 1—Schematic illustration of weld sample preparation
Schaeffler4 recognized that the chemical composit ion of the weld composite region has the greatest influence on the amount of residual ferrite. He was the first to develop a diagram which predicted the volume percent ferrite in terms of the ratio of austenite-stabilizing elements (Ni, Mn , C) to ferrite-stabilizing elements (Cr, Si, Mo, Cb). More recently, DeLong et a/.5 have revised this diagram by adding the influence of nitrogen as an austenite stabilizer. Unfortunately, although both diagrams provide a means for predicting the ferrite content from the weld chemistry, significant deviations frequently result from a combination of process-related factors.
Several investigators67 have reported that the as-welded ferrite morphology is strongly dependent on the welding parameters. In general, if the weld heat input is high, the solidification substructure tends to be coarsened and results in a more widely spaced ferrite network. Lower heat inputs are accompanied by faster cooling rates and promote the formation of finer substructures which provide a finer ferrite network. In addit ion, since the local cooling rate varies in a continuous fashion from the fusion line to the weld centerline, point-to-point variations in ferrite morphology are frequently observed with in the weldment.
Takalo et a/.8 have reported that a direct correlation exists between the
morphology and the relative amount of ferrite present in the as-welded microstructure. They propose that, as the percentage of delta ferrite increases in the range from 5 to 15 vol-%, the vermicular morphology is gradually replaced by " lathy" ferrite. However, solidification studies performed by both Matsuda3 and Fredriksson- with 18Cr-8Ni stainless steels indicated that, upon rapid quenching from the solidification range, the as-welded vermicular structure was replaced by the same lath-like morphology observed by Takalo. Thus, it is apparent that the cooling rate of the weldment exerts considerable influence upon the final ferrite morphology.
Wi th the aid of both the solidification model and the proposed mechanism for solid-state transformation which were developed in an earlier paper,1 it is now possible to explain the origin of the wide range of duplex microstructures which are observed in austenitic stainless steel weldments.
Objectives
The objectives of this investigation are threefold:
1. To produce a series of as-welded microstructures which reflect the range of ferrite morphologies commonly observed in the composite region of austenitic stainless steel weldments.
2. To explain these microstructures with the aid of the solidification mod-
Pulsed-GMAW
Average current: Average voltage: Filler metal feed
rate: Carriage travel: Shielding gas:
Pulse rate:
310/304 and 312/ 304L
100 ± 5 A 27 ± 1 V, 150 ipm
16 ipm 40 cfh of
2% O., 120 pps
GTAW (di lut ion-control pass):
Voltage, V: Travel speed, ipm: Shielding gas:
Current, A
310/304 10 ± 0.5 8 40 cfh
argon 80-130
dcrp
argon wi th
312I304L 12 ± 0.5 6 40 cfh
argon 150-200
el proposed in an earlier report.1
3. To discuss the importance of the solid-state transformation of ferrite to austenite in determining the morphology and distribution of the retained ferrite.
Experimental Procedure
In order to produce weld composite regions with the desired ferrite distribution and morphology, two sets of welds were made using the pulsed-C M A W process. In one case, a wholly austenitic Type 310 stainless steel filler metal was deposited on a Type 304 sheet, whi le in the other case, a highly-ferritic Type 312 filler metal was deposited on a Type 304L sheet. In both cases, 0.045 in. (1.1 mm) diameter filler metal wire and 0.1 in (2.5 mm) thick sheet were employed. After welding, the weld reinforcement was removed by machining; the portion of the weld composite region which remained, together with a controlled amount of the surrounding base metal, was melted using the GTAW process. This three-step procedure is illustrated schematically in Fig. 1.
By varying the heat input of the final GTAW weld pass, it was possible to control the degree to which the initial composite region was diluted by the base material, thereby producing a series of weldments with various ratios of ferritizers to austenitizers. A uniform heat sink was ensured by clamping the samples in a massive, copper welding fixture during'the final GTAW pass. The welding parameters for both the pulsed-GMAW and GTAW processes are listed in Table 1.
Both the chemical composition of the alloy weldments, as calculated from di lut ion measurements, and the ferrite number (F. N.), as measured by the Magne Gage, are listed in Table 2.
Electron beam microanalysis of se-
W E L D I N G RESEARCH SUPPLEMENT I 49-s
Table 2—Weldment Di lut ion Measurements, Ferrite Numbers and Chemical Compositions
Weldment
310/304
312/304L
Di lut ion
25 40 60 65 75
70 65 60
Ferrite number'
0 <1.0
1.5 3.0 4.0
Chemical composit ions, %'"
Cr
24.9 23.6 21.8 21.4 20.4
Ni
18.2 16.3 13.8 13.1 11.8
Si
0.5 0.5 0.55 0.55 0.55
1.6 1.6 1.6 1.6 1.6
C
.095
.088
.077
.074
.068
.018
.020
.024
.025
.027
.005
.007
.009
.010
.012
304
304 L
Autogenous
Autogenous
4.5
6.0
18.3
18.6
8.55
9.61
0.6
0.6
1.55
1.5
0.055
0.027
.03
.005
.015
.005
16.0 20.0 30.0
21.8 22.3 22.8
9.4 9.3 9.2
0.6 0.6 0.6
1.5 1.5 1.5
0.04 0.05 0.06
.03
.03
.03
.007
.007
.008
"As measured by the Magne Gage. "Calculated from weldment dilution.
lec ted reg ions o f t he a l loy w e l d m e n t s was p e r f o r m e d us ing a p o i n t c o u n t t e c h n i q u e . A n o m i n a l spec imen current of 0.25 /xA was m a i n t a i n e d at an acce le ra t ing vo l tage o f 20 kV. The spat ia l r eso lu t i on o f the e lec t ron beam was o n the o rde r of 2 -3 m ic rons .
Results O b t a i n e d W i t h 3 1 0 / 3 0 4
W e l d m e n t s
D i l u t i o n o f t h e a l l oy - r i ch Type 310 f i l ler meta l (27Cr -21Ni ) by t he Type 304 base mater ia l ( 18Cr -8N i ) resu l ted in a series o f a l loys w h i c h w e r e p r o gressively m o r e fer r i t i c in na ture . The l o c a t i o n of t h e 310/304 a l loys re lat ive to the Fe-Cr-Ni l i qu i dus and so l idus bounda r i es is i l l us t ra ted in Fig. 2. N o t e
that t he c o m p o s i t i o n o f t h e Type 304 base mater ia l places it in t h e l e f t - hand p o r t i o n o f t he d iagram w h e r e so l i d i f i ca t i on w o u l d be e x p e c t e d to occu r as del ta fe r r i te . Converse ly , t he c o m p o s i t i on o f t h e T y p e 310 f i l ler me ta l lies to t he r igh t o f t h e l i qu idus b o u n d a r y and w o u l d be expec ted to so l id i f y as p r i mary aus ten i te .
Three o f t h e a l loy w e l d pads—those w i t h 25 ,40 a n d 60% d i l u t i o n - l i e o n t h e N i - r i ch s ide o f t h e l i qu i dus b o u n d a r y bu t w i t h i n t h e borders o f t he so l idus bounda ry . As a resul t , so l i d i f i ca t i on o f these a l loys w o u l d occu r as aus ten i te a l t h o u g h , d u r i n g t he t e r m i n a l t rans ien t p e r i o d , a d i v o r c e d eu tec t i c m i x t u r e o f aus ten i te and fer r i te f o rms . By c o m p a r
i son , w e l d pads w i t h 65 and 75% d i l u t i on l ie o n the Cr - r i ch side o f t he l i qu idus b o u n d a r y and w o u l d so l i d i f y as de l ta fe r r i te . In th is case, d u r i n g t he t e rm ina l t rans ient stage o f so l i d i f i ca t i o n aus ten i te f o rms at the subgra in bounda r i es as a d i v o r c e d eu tec t i c .
By re fe rence to t he schemat i c Fe-Cr-Ni p s e u d o - b i n a r y d iag ram in Fig. 3 it is also poss ib le t o p red i c t t h e behav io r o f a w i d e range o f Fe-Cr-Ni al loys u p o n c o o l i n g f r o m t h e so l i d i f i ca t i on range. A l l oys in reg ion 1 o f th is d iagram c o r r e s p o n d t o c o m p o s i t i o n s on the N i - r i c h side of t he l i qu i dus b o u n d a r y in Fig. 2, w h i l e t he Cr - r i ch al loys in Fig. 2 c o r r e s p o n d to c o m p o s i t ions w h i c h are f o u n d in reg ion 2 o f
30
x o
& 20 -
10
5
e / 304 /
1
1 1
J " .
1 /
/ / / / •/ k/: y\/\v
f
/ T
/ s 310
N 25%
S 4 0 %
- 60 %
- 6 5 %
- 7 5 %
LIQUIDUS
i 10 20
WT % NICKEL
Fig. 2—Location of the 310/304 variable dilution weldments relative to the Fe-Cr-Ni ternary liquidus and solidus boundaries
•PRIMARY AUSTENITE - -PRIMARY DELTA FERRITE-
FE-CR-NI PSEUDO- BINARY
CHROMIUM - NICKEL
Fig. 3—Schematic pseudo-binary diagram of the Fe-Cr-Ni ternary system illustrating the effect of composition on austenite-ferrite morphology in austenitic stainless s_teel weldments
5 0 - s l F E B R U A R Y 1 9 8 0
W:fwA
i i . .. -..
Fig. 4—Transverse section of a 310/304 weldment with 25% dilution. Note dark-etching unmixed zone separating the whol-ly-austenitic composite region (right) from the Type 304 base metal (left). Mixed acid etch, X100
the pseudo-binary diagram. Under equil ibrium conditions, delta
ferrite formed during the solidification of alloys in regions 1 and 2 of Fig. 3 would experience a diffusion-controlled transformation to austenite upon cooling to room temperature. However, as a result of the rapid cooling rates experienced in weldments, a diffusion-controlled, nucleation and growth reaction is unlikely and the majority of the delta ferrite must transform to austenite via a composit ion-invariant, massive reaction. Only that portion of the ferrite which was sufficiently enriched in chromium and depleted in nickel remains stable, at room temperature. (The reader is referred to an earlier report1 in which the mechanics of both the solidification process and the solid-state transformation were rigorously reviewed.)
Effect of Minor Alloying Elements
The silicon and manganese contents of the series of alloys listed in Table 2 were such as to have nearly equal and opposite effects on the stability of ferrite and austenite, respectively, and so should not significantly influence the solidification behavior of the alloys. However, since carbon is a powerful austenite former, the presence of carbon in the melt prior to solidification should tend to shift the position of the liquidus and solidus boundaries to the left in Fig. 2 (to the right in Fig. 3) and promote the format ion of austenite as the primary phase of solidification. Increased di lut ion of the Type 310 filler metal by the Type 304 base material mitigates this effect and as a result the solidification behavior of the more dilute weld pads should be more closely approximated by Fig. 2.
2
2 O en
<fs I -5
30
28 -
26
24
22
20
PRIMARY AUSTENITE SOL I D l F I C A T I O N
• —•—• .—-•^•" V NICKEL
CHROMIUM «'
Subgrain core
/°'N ^ o
subgrain boundary
'O—o
subgrain core
20
18
16
14
12
5 10 15 20 25
DISTANCE (MICRONS) Fig. 5—Electron probe microanalysis trace across a subgrain boundary in the fully austenitic composite region in Fig. 4
Fully Austenitic 310/304 Weldments Fusion Zone Microstructure. Type
310/304 weldments wi th 25% di lut ion solidified as primary austenite and exhibited a fully austenitic, cellular dendritic substructure. The transverse weld section, shown in Fig. 4, indicates that the austenite subgrains wi th in the composite region to the right of the arrows show no evidence of retained ferrite. However, attention is drawn to the band to the left of the composite region which contains a vermicular network of dark-etching delta ferrite in an austenite matrix.
Upon closer examination it was discovered that the delta ferrite formed over much of this region. This was the result of the melting and resolidifica-t ion of the stagnant boundary layer which forms when the dissipation of the superheat from the weld pool causes in-situ melting of a small port ion of the adjacent base metal. Since the composit ion of this "unmixed zone" corresponds to that of the Type 304 base metal, solidification occurs as primary delta ferrite and the retained ferrite evident in Fig. 4 is distributed along the cores of the subgrains. Note that ferrite is also present along segregate bands in the heat-affected zone to the left of this region.
Microanalysis of the Fully Austenitic Substructure. According to theory, the solidification product formed during
the initial transient period of Type 310, when austenite is the primary phase, is depleted in chromium and slightly enriched in nickel. Following this initial transient period, solidification occurs under steady state conditions whereby austenite of near-nominal composition forms from the liquid until the onset of the terminal transient. At this point, the last remaining liquid in the cellular dendrite interstices solidifies to form a divorced eutectic mixture of austenite and ferrite.
According to theory, the last remaining liquid should be enriched in both Cr and Ni as it approaches the ternary eutectic composition (49Cr-43Ni-8Fe). This redistribution of Cr and Ni is supported by the concentration profiles shown in Fig. 5. Note the depletion in Cr at the cores of the cells and the enrichment in both Ni and Cr at the cellular dendrite boundaries.
Unfortunately, the spatial resolution of the electron beam microprobe (2-3 microns) is so large compared to both the width of the initial and final transient regions that only trends in the compositional variation across a sub-grain can be identified in Fig. 5. Chromium depletion at the cell core and enrichment of both Ni and Cr at the subgrain boundary is evident. However, the relatively flat nickel profile indicates that the distribution
W E L D I N G RESEARCH SUPPLEMENT I 51-s
Fig. 6—Transverse section of a 310/304 weldment with 40% dilution. Mixed acid etch, X100 (reduced 38% on reproduction)
• • • • •
i-r
• • I 4 0 . . m I
Fig. 7—Portion of the composite zone microstructure in Fig. 6. Note the scattered particles of dark-etching ferrite along the subgrain boundaries. Mixed acid etch, X500 (reduced 38% on reproduction)
coe f f i c ien t for n icke l d u r i n g p r ima ry aus ten i te so l i d i f i ca t i on is c lose to u n i ty. Thus, except for t h e t e r m i n a l t r an sient reg ion t h e en t i re s t ruc tu re c o n tains n e a r - n o m i n a l a m o u n t s o f n i ck el .
From the m i c r o s t r u c t u r e s h o w n in Fig. 4 and the data in Fig. 5, it is o n l y poss ib le t o c o n c l u d e tha t t h e c h r o m i u m e n r i c h m e n t o f t he cel l b o u n d a r i e s was insu f f i c ien t to p r o m o t e t he f o r m a t i o n of a reso lvab le v o l u m e o f fer r i te .
310/304 Weldments wi th F.N. < 1.0
Fusion Zone Microstructure. I n creased d i l u t i o n of t h e a l l oy - r i ch w e l d nugget by the a l l oy - lean Type 304 base meta l p r o d u c e d w e l d m e n t s w h i c h c o n t a i n e d a greater percen tage o f re ta ined de l ta fer r i te . A c o m p o s i t e w e l d m e n t w i t h 40% d i l u t i o n w h i c h e x h i b i t e d less t h a n F.N. 1 is s h o w n in Fig. 6. In th is case, so l i d i f i ca t i on o c c u r r e d as re la t ive ly coarse aus ten i te subgra ins, a l t h o u g h c loser exam ina t i o n of t h e subgra in bounda r i es in Fig. 7 ind icates tha t occas iona l l y t he d i v o r c e d eu tec t i c fe r r i te w h i c h f o r m e d f r o m the last l i qu i d t o so l id i f y
30
2 28
g 26
24 -eN t -
5 22
20 -
PRIMARY A U S T E N I T E SOLIDIF ICATION
•—• -•" ^ • -NICKEL
- •—•—• -
CHROMIUM
. /
subgrain core
\y . • N \ _
subgrain boundary
J _
subgrain boundary
J . - L
20
Ifl
lb
14
12
10
S - I
o\
z o * m i -
5 10 15 20 25
DISTANCE (MICRONS) Fig. 8—Electron probe microanalysis trace across two ferrite-free su boundaries in Fig. 7
r ema ined stable at r o o m t e m p e r a t u r e . A d u p l e x u n m i x e d z o n e is also ev i
den t in Fig. 6. H o w e v e r , it does no t appear as ex tens ive as tha t observed in Fig. 4, p r o b a b l y as a result o f e i ther t he more e f f i c ien t m i x i n g o f t he h igher heat i n p u t GTA w e l d pass used to ach ieve f ina l d i l u t i o n or the smal ler d i f f e rence b e t w e e n the c o m p o s i t i o n o f the c o m p o s i t e z o n e a n d that o f t he base me ta l .
Microanalysis of the Composite Zone Substructure. A c o n c e n t r a t i o n p ro f i le across t w o c o m p o s i t e z o n e ce l l bounda r ies w h i c h w e r e d e v o i d o f ferr i te is s h o w n in Fig. 8. Since so l id i f i ca t i o n o f th is a l l oy w e l d m e n t also occurs as p r imary aus ten i te , t he cel l b o u n d aries are en r i ched in c h r o m i u m relat ive to the s u r r o u n d i n g subs t ruc tu re . H o w e v e r , d u r i n g t he f ina l t rans ient pe r i od m o r e de l ta fe r r i te is f o r m e d than in t he c o m p o s i t e reg ion s h o w n in Fig. 4. As a resul t , su f f i c ien t eu tec t i c ferr i te is re ta ined to be i d e n t i f i a b l e at r o o m t e m p e r a t u r e .
W i t h re fe rence to t he p s e u d o - b i n a r y d iagram in Fig. 3, it is ev i den t tha t , as the rat io o f c h r o m i u m to n i cke l increases w i t h i n reg ion 1 , t he a m o u n t of eu tec t i c fer r i te w h i c h fo rms as a so l i d i f i ca t i on p r o d u c t s h o u l d increase p r o p o r t i o n a t e l y .
310/304 Weldments wi th F.N. 1
Fusion Zone Microstructure. W e l d ments w h i c h we re d i l u t e d 60% w i t h Type 304 base meta l e x h i b i t e d F.N. 1 in
t he a s - w e l d e d s t ruc tu re w h e n tested w i t h the M a g n e Gage. Figure 9 is a transverse sec t ion of t he c o m p o s i t e z o n e in th is t ype o f w e l d m e n t . It reveals t ha t a g rada t i on in t he a m o u n t of re ta ined fer r i te exists.
W i t h i n t he c o m p o s i t e zone , de l ta fe r r i te is present in p r o p o r t i o n a t e l y greater a m o u n t s near t he in te r face o f the u n m i x e d z o n e as a result o f t h e d i f f us i on g rad ien t w h i c h exists bet w e e n the u n m i x e d z o n e and the c o m pos i te z o n e . H o w e v e r , it is s ign i f i can t that all the re ta ined fer r i te was at aus ten i t i c subgra in bounda r i es . T h e r e fore , the a l loy must have so l i d i f i ed as p r imary aus ten i te , as p red i c t ed by Fig. 2.
Microanalysis of the Composite Zone Substtucture. E lec t ron beam m i croanalys is was p e r f o r m e d in t he c o m posi te z o n e reg ion w h i c h e x h i b i t e d scat tered eu tec t i c fe r r i te a l ong cel l bounda r i es o f t he p r ima ry aus ten i t e subgrains as i l lus t ra ted by Fig. 10. A n a l ysis of reg ion A -A , w h i c h in te rsec ted t w o bounda r i es c o n t a i n i n g fe r r i te , i nd i ca ted that c h r o m i u m e n r i c h m e n t and n icke l d e p l e t i o n occurs w h e r e ferrite is present , as i l lus t ra ted by t he c o n c e n t r a t i o n p ro f i l e in Fig. 1 1 . In c o n trast to the fu l l y aus ten i t i c s t ruc tures , the presence o f fe r r i te at t he subgra in bounda r i es results in a local decrease in t he n i cke l c o n t e n t as the beam is cen te red o n the fe r r i t i c phase. The f o r m a t i o n of t he d i v o r c e d fe r r i te in t he fe r r i t e -aus ten i te eu tec t i c f r o m the last
52-sl FEBRUARY 1980
.
Fig. 9—Transverse section of a 310/304 weldment with 60% dilution which exhibited F.N. 1.'Mixed acid etch, X100 (reduced 38% on reproduction)
X '£...
2
2 o cr X o
o\ ( -5
30
28
26
24
22
20
PRIMARY A U S T E N I T E S O L I D I F I C A T I O N
~- V ^ \
ferr i te
CH ROMIUM 0 ^ ° \ 7 N>_o^
r*-+ •
NICKEL "
14
12
10
- 1
ON
O
m
o — o subgrai n boundary
J L
Subgrain boundary
1 - L
16
Fig-
5 10 15 20 25
DISTANCE (MICRONS) 11—Electron probe microanalysis trace across region A-A in Fig. 10
I 4 0 ,,m I . . ' —* " ~ • » _
Fig. 10— Portion of the composite zone microstructure in Fig. 9. Ferrite particles are located at subgrain boundaries. Mixed acid etch, X500 (reduced 38% on reproduction)
l i q u i d t o so l i d i f y causes a re jec t i on o f n i cke l f r o m the fer r i te in t he b o u n d a r y reg ion .
Theory p red ic ts tha t t he c o m p o s i t i o n o f t he l i q u i d in t he t e r m i n a l t ran sient must a p p r o a c h the te rnary eu tec t ic c o m p o s i t i o n (49Cr -43N i -8Fe ) . C o n sequen t l y , t h e f ina l fe r r i te w h i c h fo rms at t he subgra in b o u n d a r y mus t be h igh ly e n r i c h e d in c h r o m i u m w h i l e t he s u r r o u n d i n g aus ten i te is e n r i c h e d in n i cke l . H o w e v e r , t he t e r m i n a l t r an sient represents such a smal l f r ac t i on o f t he b o u n d a r y v o l u m e tha t it is n o t reso lvable by t he e l ec t r on m i c r o p r o b e . H o w e v e r , t he increase in c h r o m i u m observed w h e n t h e beam is cen te red o n t h e fe r r i te at t he subgra in b o u n d a r y in Fig. 11 is cons is ten t w i t h w h a t t he t heo ry pred ic ts .
310/304 Weldments wi th F.N. 3
Fusion Zone Microstructure. A n increase in t he d i l u t i o n by t he base meta l f r o m 60 to 65% s h o u l d p lace t h e n o m i n a l c o m p o s i t i o n o f t h e c o m p o s i te z o n e in t he p r imary de l ta fe r r i te
reg ion o f the Fe-Cr-Ni d i a g r a m , as s h o w n in Fig. 2. As a resul t th is a l l oy so l i d i f i ed as p r imary de l ta fe r r i te and e x h i b i t e d F.N. 2.5-3.5 in t h e as -we lded m ic ros t r uc tu re . The m a r k e d increase in t he a m o u n t o f re ta ined fe r r i te in th is a l loy w e l d m e n t is ev i den t f r o m the transverse sec t ion o f t he w e l d s h o w n at X100 in Fig. 12. The u n m i x e d z o n e is n o t as appa ren t as in Figs. 4, 6, and 9 because of t h e s imi la r i t y in t h e c o m pos i t i ons o f t h e c o m p o s i t e z o n e and t h e base me ta l .
That in i t ia l so l i d i f i ca t i on o f th is a l loy o c c u r r e d as de l ta fe r r i te is s u p p o r t e d by the presence o f re ta ined de l ta ferr i te at t he cores of t he subgra ins. H o w ever, t h e presence o f some re ta ined fer r i te at t he subgra in b o u n d a r i e s suggests tha t a m i x e d m o d e of so l i d i f i ca t i o n was expe r i enced . It is pos tu l a ted tha t , d u r i n g so l i d i f i ca t i on as fe r r i te , segregat ion of n icke l in f r o n t of t h e advanc ing s o l i d - l i q u i d in te r face and at subgra in bounda r ies caused a shi f t t o so l i d i f i ca t i on as aus ten i te . Since t h e n o m i n a l c o m p o s i t i o n o f th is a l loy is just t o t h e left o f t he m i n i m u m l i q u idus in Fig. 2, a s l ight e n r i c h m e n t o f t h e l i q u i d in N i c o u l d cause such a t rans i t i on t o so l i d i f i ca t i on as p r imary aus ten i te . A s imi lar shi f t in so l i d i f i ca t i o n m o d e f r o m p r imary de l ta fe r r i te t o p r imary aus ten i te as a resul t o f t h e segregat ion o f n icke l in f r o n t o f t h e s o l i d - l i q u i d in te r face has been rep o r t e d p rev ious l y by Fredr iksson.2
Microanalysis of Fusion Zone Mi-
/ • ,
. • • • . - : . - ' ••••"
0 . 2mm
Fig. 12— Transverse section of a 310/304 weldment with 65% dilution which exhibited F.N. 3. Mixed acid etch, X100 (reduced 38% on reproduction)
crostructures. The presence of reta ined fe r r i te at t he cores o f t he f o r m e r p r imary de l ta fe r r i te subgra ins is d o c u m e n t e d in Fig. 13. The c o n c e n t r a t i o n p ro f i le s h o w n in Fig. 14, t raversed reg ion 3-3, w h i c h encompassed t w o p r imary dend r i t es a n d the i n t e r d e n -dr i t i c b o u n d a r y separat ing these d e n dr i tes. These data i nd i ca te tha t t h e cel l cores are en r i ched in c h r o m i u m and d e p l e t e d in n i cke l .
It is in te res t ing to n o t e , h o w e v e r , tha t a smal l increase in c h r o m i u m c o n cen t ra t i on also occurs at t he subgra in b o u n d a r y m i d w a y b e t w e e n the cel l cores. So l i d i f i ca t i on t h e o r y p red ic ts tha t t he c h r o m i u m shou ld decrease in
W E L D I N G R E S E A R C H S U P P L E M E N T I 53 -s
this region if the primary phase is delta ferrite. Thus, it seems likely that the mode of solidification shifts from one of primary delta ferrite to one of primary austenite during the terminal transient stage. This change in solidification mode would cause chromium to be rejected to the l iquid and would result in an increase in the amount of eutectic ferrite formed during the terminal transient. Careful inspection of Fig. 13 indicates that some particles of divorced eutectic ferrite formed along interdendritic boundaries, as denoted by the arrows.
The nominal composition of this alloy lies close to the composit ion separating primary delta ferrite from primary austenite solidification. Because of this, localized fluctuations in the composit ion of the l iquid ahead of the advancing interface which lower the Cr/Ni ratio can easily modify the solidification mode to produce substantial amounts of primary austenite. The microstructure presented in Fig. 15 can be explained by such a change in solidification mode.
Similar microstructures were observed frequently in transverse weld sections in this alloy series. Retained ferrite at the subgrain boundaries, indicated by the arrows in Fig. 15, is believed to have formed in this way. Such a distribution of delta ferrite suggests that enrichment of the liquid in chromium during the portion of the solidification process when austenite served as the primary phase, promoted a more extensive divorced eutectic reaction and resulted in a greater percentage of eutectic ferrite in the sub-grain boundaries.
Figure 16, which summarizes the electron beam microanalysis data for region 1-1 in Fig. 15 indicates a 4-6 wt-% increase in the concentration of chromium and a 3-5 wt-% decrease in
• 4 ' . ; / '
nickel in the intercellular region relative to the concentration present at the core of the subgrains.
This represents a substantial chromium enrichment and nickel depletion relative to the microprobe data for the ferrite in Fig. 14 and offers further proof that the ferrite in Fig. 15 is a divorced eutectic constitutent retained at the primary austenite sub-grain boundaries.
310/304 Weldments with F.N. 4
Fusion Zone Microstructures. We ld ments which were di luted 75% by the final GTA pass and autogeneous single pass GTA welds on Type 304 sheet exhibited about the same amount of delta ferrite in the fusion zone. Since both series of fusion zone compositions lie wi th in the primary delta ferrite region of the ternary diagram in Fig. 2, initial solidification occurred as primary ferrite in both cases.
The transverse weld section of the 75% diluted alloy is shown at X100 in Fig. 17. In addition to the fusion zone ferrite, the higher GTA energy input required to produce 75% dilut ion increased the width of the partially-melted zone where retained delta ferrite is visible. However, the composition of the base metal and the composite zone are so similar that no unmixed zone can be identified wi th the etching procedure used.
The distribution of delta ferrite in
the fusion zone of both the alloy wi th 75% di lut ion and the autogenous Type 304 weldment appeared similar and was generally observed to be along the cores of the cellular dendrit ic primary delta ferrite subgrains. The vermicular, semicontinuous ferrite morphology in a transverse weld section is shown at X500 in Fig. 18. Electron beam micro-analysis across primary ferrite den-drites wi th retained delta ferrite at the core produced concentration profiles similar to those observed in Fig. 14, and indicated that solidification occurred as delta ferrite.
Results Obtained wi th 312/304L Weldments
The compositions of the 312/304L composite welds (Table 2) are such that the primary phase during solidif ication is always delta ferrite. Although the solidification behavior of the four alloys is similar, the particular composition controls both the mechanism of solid-state transformation of ferrite to austenite and the morphology of the ferrite in the as-welded microstructure. Consequently, the specific alloy compositions wil l be discussed wi th reference to the Fe-Cr-Ni pseudo-binary diagram in Fig. 3.
Autogenous Type 304L Weldment
The fusion zone microstructure of an autogenous Type 304L weldment
Fig. 13—Portion of the composite zone microstructure in Fig. 12 which solidified as primary delta ferrite. Arrows indicate particles of divorced eutectic ferrite along the subgrain boundaries. Mixed acid etch, X500 (reduced 38% on reproduction)
2 3
2 O or x o
N? oN
S
2 8
2 6
2 4
22
2 0
18
m
fer
•
\ 1
r i te
/
A / o
y
subq roi core
- i -
P R I M A R Y D E L T A F E R R I T E
S O L I D I F I C A T I O N
f e r r i t e
N I C K E L
. / # ' ' • — - — . ^
7 \ •
C H R O M I U M
v \ o ^ - C
/ 3
n s u b g r a i n s u b g r a i n boundary core
« — J 1 1
18
16 *
ON
14 O ~*.
12 m r~
10
- 8
5 10 15 20 25
DISTANCE (MICRONS)
Fig. 14— Electron probe microanalysis trace across region 3-3 in Fig. 13
54-sl FEBRUARY 1980
* ^ 1' /
ir-i>t
40
<m^w:'*wm Fig. 15—Portion of the composite zone microstucture in Fig. 12 which solidified as primary austenite. Dark-etching ferrite stringers appear along subgrain boundaries. Mixed acid etch, X500 (reduced 38% on reproduction)
exhibited a vermicular morphology of delta ferrite in an austenite matrix (F.N. 6), as shown in Fig. 19A. The nearly continuous networks of delta ferrite correspond to the Cr-enriched, Ni-depleted regions formed during the initial transient period and are thus located at the cores of the cellular dendritic subgrains. The nominal composition of Type 304L places it near the midpoint of region 2 in the pseudo-binary diagram in Fig. 3. The composition of the initial solid can be estimated from the intercept of an isotherm at the liquidus for the nominal composition with the solidus curve in Fig. 3. Thus the first solid to form at the cores of the subgrains in Type 304L lies to the right of region 4 of the pseudo-binary diagram.
The majority of the material in the fusion zone of Type 304L solidifies as ferrite of approximately nominal composition during the steady-state stage of solidification. Once solidified, it enters the two-phase austenite plus ferrite field almost immediately and transforms to austenite by a massive-type transformation. However, the Cr-enriched, Ni-depleted dendrite core whose composit ion lies to the right of region 4 is essentially a ferritic stainless steel and, therefore, remains stable at room temperature. Since the cooling rate of the weldment is relatively fast, redistribution of solute by diffusion is minimal and the final as-welded microstructure consists of a vermicular morphology of the delta ferrite formed during the initial transient in a matrix of austenite as shown in Fig. 19A.
Microanalysis of this structure indicated that the dendrite core is, as expected, enriched in chromium and depleted in nickel. Also, except for the subgrain boundaries, the remainder of the alloy consists of austenite of near-nominal composit ion. In addit ion, high resolution STEM microanalysis
2 r> 2 o
ON
t -
28 -
26 -
24 -
22 -
20
PRIMARY AUSTENITE SOLIDIFICATION
ferr i te f e r r i t e
,•—
subgrai n boundary
subgrain boundary
J _ X
16
14
\i>
10
8
6
S H
\° ON
Z
o 7^ m r~
5 10 15 20 25
DISTANCE (MICRONS) Fig. 76—Electron probe microanalysis trace across region 1-1 in Fig. 15
performed by Lyman et al9 on an alloy of similar composit ion revealed that the composit ion of the retained ferrite approximates that predicted by the solidification model.1
312/304L Weldment with F.N. 16
By enriching the composit ion of the Type 304L base metal wi th a ferritic Type 312 filler metal (29.5 Cr-9 Ni), it was possible to produce alloy composite regions with increased amounts of delta ferrite. The duplex microstructure in Fig. 19B exhibits a continuous ferrite network of F.N. 16.
Since the nominal composition of the composite zone of this alloy would lie in the right-hand portion of region 2 of the schematic pseudo-binary diagram of Fig. 3, the initial solid at the dendrite core is more enriched in chromium than the corresponding region in the autogenous Type 304L weldment and lies on the solidus curve still farther to the right of region 4. As a result, a greater proportion of the initial transient region in this alloy wil l cool through the high-Cr, low-Ni single phase region to the right of the ferrite solvus. However, thefaulk of the alloy solidifies as delta ferrite of nominal composit ion, and only a portion of the alloy solidified during the initial transient stage is retained as delta ferrite.
Upon cooling from the solidification range, the ferrite of near nominal composition transforms to austenite by a
massive transformation. However, the dendrite core region whose composition lies to the right of region 4 persists as ferrite to room temperature. A sufficient volume of the material formed during the initial transient has a Cr/Ni ratio high enough to form continuous ferrite networks along primary and secondary dendrite cores. The resulting gray-etching, blocky ferrite and the lacy, dark-etching networks in Fig. 19B represent these primary dendrite cores.
Microanalysis of the structure shown in Fig. 19B indicated that the austenitic regions exhibited near nominal composition. Profiles taken across the retained ferrite (in those cases where the particles were of sufficient size to obtain adequate spatial resolution) indicated that a smoothing of the initial concentration gradients wi th in the ferrite had occurred. This undoubtedly results from the large temperature range over which the composition at the core remains within the single phase ferrite region and also reflects the fact that diffusion is more rapid in ferrite than in austenite.1"
312/304L Weldments with F.N. 20
As the composition region becomes more enriched in chromium, the nominal composit ion of the weldment shifts into region 3 of the pseudo-binary diagram in Fig. 3. As shown in Fig. 19C, the resulting weld micro-structure exhibits relatively large re-
W E L D I N G RESEARCH SUPPLEMENT I 55-s
Fig. 17—Transverse section of an autogenous weld on Type 304 with F.N. 4. Mixed acid etch, XIOO (reduced 38% on reproduction)
c - . ' <• Fig. 18—Portion of the fusion zone micro-structure in Fig. 17. Dark-etching ferrite is located along the cores of the original primary delta ferrite subgrains. Mixed acid etch, X500 (reduced 38% on reproduction)
gions of acicular austenite in a ferrite matrix surrounded by a nearly cont inuous austenite network at the grain boundaries. It appears that the transformation to austenite, which nucleates at the grain boundaries, occurs along specific habit planes in the ferrite; this ultimately results in the formation of alternate austenite and ferrite laths.
Massalski et a/.11 have reported a similar microstructural transition as a function of both composit ion and cooling rate in the Fe-Ni system. They found that, as either the solute concentration or the cooling rate at a given concentration was increased, the thermally-activated massive transformation was suppressed and decomposition on cooling occurred by a mar-tensitic process. The microstructure they observed consisted of parallel laths along specific habit planes and suggests that a similar transformation may be responsible for the austenite-ferrite morphology in Fig. 19C.
Microanalysis of the lath-like regions in the F.N. 20 weldment indicated that a relatively flat concentra-
Fig.19— Duplex stainless steel weld metal microstructures: A—discontinuous vermicular ferrite network, F.N. 6; B—continuous vermicular morphology, F.N. 16, C— acicular austenite-ferrite morphology surrounded by a continuous austenite network, F.N. 20; D—Widmanstat-ten austenite in a ferrite matrix, F.N. 30. Ferrite is the dark-etching phase. Mixed acid etch, X500 (reduced 38% on reproduction)
tion profile exists wi th in the individual grains and that the composit ion of these regions approximates the nominal composit ion of the composite zone. Unfortunately, the spatial resolution of the electron probe is so large compared to the width of the individual laths that it is impossible to determine the composition of the individual phases. However, high resolution STEM microanalysis performed by Ly-man et al.1- on a similar acicular microstructure indicated (as would be expected in a martensitic transformation) that little difference in composition exists between phases in this lathlike microstructure.
The transition from the vermicular ferrite network observed in Figs. 19A and 19B to the acicular morphology present in Fig. 19C can be explained with the aid of the pseudo-binary diagram. Wi th in region 3 (Fig. 3) the primary delta ferrite must cool through a substantial temperature interval before encountering the ferrite solvus. Since the diffusion rates for chromium and nickel are rapid wi th in this region between the ferrite solidus and the ferrite solvus, the compositional gradients produced during solidification are "smoothed" prior to the onset of the solid-state transformation. Howev
er, it is possible that the cooling rate through the austenite plus ferrite region is still sufficiently rapid in fusion welds to cause a composit ion-invariant transformation of ferrite to austenite instead of a diffusion limited transformation.
Since the diffusion rate decreases and the strain energy accompanying the ferrite-to-austenite transformation increases as the temperature decreases, a massive transformation by migration of a planar interface is impossible below the austenite solvus in region 3. Instead, the transformation occurs by the formation of parallel laths wi th in the prior delta ferrite grains and results in the acicular morphology shown in Fig. 19C.
312/304L Weldments with F.N. 30
If the Cr/Ni ratio of the composite fusion zone is increased sufficiently, the nominal composit ion of the alloy wil l fall wi th in the bounds of region 4 in Fig. 3. Alloys in this composit ion range exhibit still another austenite-ferrite morphology.
The fusion zone microstructure of a 312/304L weldment wi th 60% di lut ion which exhibited a F.N. 30 is illustrated in Fig. 19D. It should be noted that the
56-s l FEBRUARY 1980
grain size is considerably larger than that of the other alloys and that a continuous, light-etching austenite network exists at the prior ferrite grain boundaries. In this alloy, partial transformation to austenite occurs by growth of nuclei formed at the grain boundaries along habit planes in the gray-etching, ferrite matrix to form Widmanstatten platelets of l ight-etching austenite.
Microsegregation of both Cr and Ni exists wi th in the fusion zone immediately fol lowing solidification. However, diffusion is undoubtedly effective in reducing the chemical inhomo-geneity during cooling through the expanded temperature range between the solidus and solvus for alloys in region 4. Furthermore, a massive transformation is impossible for alloys in region 4 since the austenite solvus line lies to the left of this range of compositions. Thus, for alloys wi th a nominal composition in region 4, the transformation to austenite requires a modif ication in composit ion and is therefore diffusion control led.
As the nominal alloy composit ion shifts from the left-hand to the right-hand boundary of region 4 in Fig. 3, the proportion of ferrite to austenite increases until alloys to the right of region 4 would remain whol ly ferritic at room temperature.
Microanalysis of adjacent austenite and ferrite regions wi th in the fusion zone shown in Fig. 19D revealed a significant compositional difference between the two phases. This offers further proof that, wi th in region 4, transformation of ferrite to austenite requires a change in composit ion and is thus l imited by the rate of diffusion of chromium and nickel across the interphase interface.
Transition Microstructures
The pseudo-binary diagram illustrated in Fig. 3 is a schematic representation of the Fe-Cr-Ni system. Moreover, it is applicable to a wide variety of commercial austenitic stainless steel filler and base metals which range from 55 to 75 wt-% iron. This range includes Types 304, 304L, 316, 308, 309, 310, and 312.
It should be noted that the position of the alloy composit ion wi th in regions 1-4 does not preclude the formation of alternate microstructures. For example, an alloy whose composition lies near the boundary of region 2 and region 3 may exhibit both vermicular and acicular ferrite in the as-welded microstructure. Small compositional fluctuations wi th in the liquid in advance of the solid-l iquid interface may also alter the ultimate ferrite morphology by effecting a local change in the composit ion of the alloy and thus
altering the nature of solid-state, fer-rite-to-austenite transformation in that region.
Effect of We ld Cooling Rate
The rate at which the weldment cools through the two-phase austenite plus ferrite region can also have a significant effect on both the ferrite distribution and morphology. Studies at RPI" have shown that if an autogenous weld in Type 304 (which normally exhibits a vermicular as-welded microstructure) is quenched rapidly from the solidification range, an acicular microstructure wi th a greater-than-normal ferrite content wi l l replace the normal vermicular structure. It was also observed that, near the fusion line where the cooling rate was the greatest, an acicular microstructure may be produced wh ich is different morphologically from that of the microstructure in the interior of the fusion zone.11
For instance, if alloys in region 2 of Fig. 3, are rapidly cooled through the two-phase region, the ferrite becomes supersaturated wi th respect to the austenite almost instantaneously. As a result, the normally planar mode of transformation of ferrite to austenite breaks down and the reaction proceeds by growth of platelets of austenite along specific habit planes in the ferrite. The large increase in driving force imparted by the supercooling below the austenite solvus as a result of the rapid quench provides the additional energy needed to create the increased area of interphase interface.
It is significant that the low F.N. acicular microstructures produced in this fashion appear similar to the acicular microstructure presented in Fig. 19C (F.N. 20). In both cases it was observed that the initial planar interface became unstable and transformation was forced to form the fine parallel, austenite laths wi th in the ferrite matrix.
Summary
On the basis of the duplex micro-structures observed during this investigation, it is possible to define four specific compositional regions on the Fe-Cr-Ni pseudo-binary diagram (Fig. 3), each of which exhibits a characteristic ferrite morphology.
• Region 1—Alloys in this range solidify as primary austenite and may form a l imited amount of ferrite as a divorced eutectic along the intercellular boundaries. If the Cr/Ni ratio of this ferrite is sufficiently high to render it stable at room temperature, it usually exhibits a semicontinuous morphology as shown in Fig. 15.
• Region 2—Alloys in this range solidify as primary delta ferrite dendrites whose cores are highly enriched in chromium and depleted in nickel. Upon cooling through the two-phase, austenite plus ferrite region, the ferrite of nominal composition formed during steady-state solidification transforms to austenite by a composit ion-invariant, massive transformation. A portion of the ferrite at the dendrite cores is sufficiently enriched in Cr and depleted in Ni to remain stable at room temperature and is characterized by a vermicular morphology as shown in Fig. 19A. As the Cr/Ni ratio increases wi th in this region, the ferrite network becomes more continuous.
• Region 3—In alloys wi th in this range, the primary delta ferrite is stable over a relatively large temperature range and a "smooth ing" of the concentration gradients in the initial transient region by diffusion occurs during cooling from the solidus to the ferrite solvus. However, the cooling rate through the two-phase region still suppresses the diffusion-controlled transformation of the ferrite and the microstructure exhibits an acicular morphology at room temperature.
• Region 4—For alloys in this range the pseudo-binary diagram predicts that ferrite and austenite should coexist in a near-equilibrium mixture at room temperature. Since the composition of the austenite formed in alloys within region 4 differs from the nominal composit ion, a massive transformation is impossible in these alloys. Consequently, a diffusion-controlled transformation of ferrite to austenite must occur upon cooling through the two-phase region. Thus, the as-welded microstructure consists of ferrite and Widmanstatten austenite which nucleates at the austenite grain boundaries and forms along specific habit planes in the ferrite. The relatively large range of temperatures over which the ferrite is stable permits extensive coarsening of the ferrite grains prior to the transformation, thus explaining the large grain size apparent in Fig. 19D.
Local fluctuations in both the alloy composit ion and the weld cooling rate along the solidification front may lead to a change in the transformation mechanism and result in a mixed microstructure. Since the kinetics of the solid-state transformation are governed by the t ime interval in the two-phase region, the weld cooling rate has a major influence on the austenite-ferrite morphology. Slow cooling rates, such as would be encountered in elec-troslag welding, for example, would tend to favor diffusion-controlled transformation mechanisms and suppress the massive transformation for alloys in regions 2 and 3.
W E L D I N G RESEARCH SUPPLEME NT I 57-s
Acknowledgment
The au thors w o u l d l ike t o thank the A l l oy Rods D i v i s i o n , C h e m e t r o n Corp . , a D i v i s i on o f A l l e g h e n y L u d l u m Corp . , fo r t he f i nanc ia l suppo r t o f th is invest i ga t i on u n d e r the auspices of t he C h e m e t r o n Fe l l owsh ip . Special thanks are also in o rde r to Dr. W i l l i a m Baes-lack III for his assistance in spec imen p repara t i on and me ta l l og raphy .
References
1. Lippold, |. l idif ication of
C , and Savage, W. F. Austenitic Stainless
"So-Steel
Weldments—Part Welding Journal, search SuppL, pp.
2. Fredriksson, quence in an 18
1, A Proposed Mode l , " 58(12), Dec. 1979, Re-362-s to 374-s. EL, "Solidif ication Se-Stainless Steel, Investi
gated by Directional Solidif ication," Met. Trans., 3 (11), 1972, pp. 2989-2997.
3. Arata, Y., Matsuda, F., and Katayama,
S., "Fundamental Investigation on Solidif ication Behavior of Fully Austenitic and Duplex Microstructures and Effect of Ferrite on Microsegregation," Trans, of JWRI, 5 (2), 1976, pp. 35-51.
4. Schaeffler, A., "Const i tut ion Diagram for Stainless Steel Weld Metal , " Metal Progress, 56 (5), 1949, pp. 680 and 680B.
5. DeLong, W., Ostrom, C , and Szuma-chowski, E., "Measurement and Calculation of Ferrite in Stainless Steel Weld Metal , " Welding Journal, 35 (11), 1956, Research SuppL, pp. 526-s to 533-s.
6. Goodwin , G. M., Cole, N. C , and Slaughter, G. M., "A Study of Ferrite Morphology in Austenitic Stainless Steel We ld ments," Welding Journal, 51 (9), 1972, Research SuppL, pp. 425-s to 429-s.
7. Asakura, S„ Wachi , H., and Watanabe, K., "The Effect of Weld ing Condit ions on the Crack Sensitivity in Austenitic Stainless Steel Weld Metals," Trans. JWS, 3 (2), 1972, pp. 34-44.
8. Takalo, T., Suutala, N., and Moisio, T., "Inf luence of Ferrite Content on Its Mor
phology in Some Austenitic Weld Metals," Mel. Trans., 10A(4), 1979, pp. 512-514.
9. Lyman, C. E., Manning, P. E., Duquette, D.)., and Hall, E., "STEM Microanalysis of Duplex Stainless Steel Weld Metal , " Scanning Electron Microscopy, Vol. 1,1978, pp. 213-220.
10. Mohar i l , D. B„ Jin, I., and Purdy, C. R., "The Effect of Delta Ferrite Formation on the Post Solidification Homogenization of Alloy Steels," Met. Trans., 5 (1), 1974, pp. 59-63.
11. Massalski, T. B., Perepezko, |. FL, and laklousky, J., "A Microstructural Study of Massive Transformations in The Fe-Ni System," Materials 5c/. and Eng., 18 (1975), pp. 193-198.
12. Lyman, C. E., and Manning, P., unpublished research performed at Rensselaer Polytechnic Institute, 1978.
13. Lippold, J. C , Ph.D. Thesis, Rensselaer Polytechnic Institute, Troy, N.Y., 1978.
14. Baeslack 111, W. A., and Lippold, ]. C , unpublished research performed at Rensselaer Polytechnic Institute, 1978.
WRC Bulletin 248 May 1979
Allowable Axial Stress of Restrained Multi-Segment, Tapered Roof Girders
by G. C. Lee, Y. C. Chen and T. L. Hsu
In this paper allowable axial stresses of restrained tapered roof girders are developed. They are particularly useful to determine the allowable stress in the interaction equation for frames consisting of segmented tapered sections because the roof girders are generally supported adequately in the lateral direction by purlins. This study has concentrated on the inplane design of roof girders when the overall column instability, above the strong axis, governs the design.
Because of the complexity of the problems due to the many parameters, the effective length factors are presented in the form of curves for both cases of sidesway permitted and sidesway prevented.
Publication of this paper was sponsored by the WRC-SSRC Joint Subcommittee on Tapered Members of the Structural Steel Committee of the Welding Research Council.
The price of WRC Bulletin 248 is $10.50 per copy. Orders should be sent with payment to the Welding Research Council, 345 East 47th Street. Room 801, New York, NY 10017.
WRC Bulletin 249 June 1979
Review of Analytical and Experimental Techniques for Improving Structural Dynamic Models by Paul Ibanez
The purpose of this paper is to review models for using experimental data to improve structural dynamic models for pressure vessels, piping systems, and their support and restraint systems. Laboratory models and scaling laws are discussed, followed by a summary of experimental results and potential bench mark cases on actual pressure vessel systems. Computer programs are also summarized, and an attempt is made to present a state-of-the-art summary of techniques for identification of structural dynamics models from experimental data.
Publication of this paper was sponsored by the subcommittee on Dynamic Analysis of Pressure Components of the Pressure Vessel Research Committee of the Welding Research Council.
The price of WRC Bulletin 249 is $11.50 per copy. Orders should be sent with payment to the Welding Research Council, 345 East 47th St.. Room 801, New York, NY 10017.