CHAPTER 6

AUSTENITIC STAINLESS STEELS

Austenitic stainless steels represent the largest of the general groups of stainless
steels and are produced in higher tonnages than any other group. They have good
corrosion resistance in most environments. The austenitic stainless steels have
strengths equivalent to those of mild steels, approximately 210 MPa (30 ksi) mini-
mum yield strength at room temperature, and are not transformation hardenable.
Low-temperature impact properties are good for these alloys, making them useful in
cryogenic applications. Service temperatures can be up to 760°C (1400°F) or even
higher, but the strength and oxidation resistance of most of these steels are limited
at such high temperatures. Austenitic stainless steels can be strengthened signi-
ficantly by cold working. They are often used in applications requiring good atmos-
pheric or elevated temperature corrosion resistance. They are generally considered
to be weldable, if proper precautions are followed.
Elements that promote the formation of austenite, most notably nickel, are added
to these steels in large quantities (generally over 8 wt%). Other austenite-promoting
elements are C, N, and Cu. Carbon and nitrogen are strong austenite promoters, as
can be seen from the various values in nickel equivalency formulas (see Chapter 3).
Carbon is added to improve strength (creep resistance) at high temperatures. Nitrogen
is added to some alloys to improve strength, mainly at ambient and cryogenic tem-
peratures, sometimes more than doubling it. Nitrogen-strengthened alloys are usu-
ally designated with a suffix N added to their AISI 300 series designation (e.g.,
304LN). The AISI 200 series (e.g., 201) are also nitrogen strengthened and are com-
monly referred to under various trade names, such as Nitronic®.

Welding Metallurgy and Weldability of Stainless Steels, by John C. Lippold and Damian J. Kotecki
ISBN 0-471-47379-0 Copyright © 2005 John Wiley & Sons, Inc.

141
142 AUSTENITIC STAINLESS STEELS

Composition Range of Standard Austenitic Stainless Steels

• 16 to 25 wt% chromium
• 8 to 20 wt% nickel
• 1 to 2 wt% manganese
• 0.5 to 3 wt% silicon
• 0.02 to 0.08 wt% carbon (0.04 wt% designated L grades)
• 0 to 2 wt% molybdenum
• 0 to 0.15 wt% nitrogen
• 0 to 0.2 wt% titanium and niobium

Austenitic stainless steels generally have good ductility and toughness and exhibit
significant elongation during tensile loading. They are more expensive than the
martensitic and low to medium Cr ferritic grades, due to the higher alloy content of
these alloys. Despite the steel cost, they offer distinct engineering advantages, par-
ticularly with respect to formability and weldability, that often reduce the overall
cost compared to other groups of stainless steels.
Although there are a wide variety of austenitic stainless steels, the 300 series
alloys are the oldest and most commonly used. Most of these alloys are based on the
18Cr–8Ni system, with additional alloying elements or modifications to provide
unique or enhanced properties. Type 304 is the foundation of this alloy series, and
along with 304L, represents the most commonly selected austenitic grade. Type 316
substitutes approximately 2%Mo for a nearly equal amount of Cr to improve pitting
corrosion resistance.
The stabilized grades, 321 and 347, contain small additions of Ti and Nb, respec-
tively, to combine with carbon and reduce the tendency for intergranular corrosion
due to Cr-carbide precipitation. The L grades became popular in the 1960s and
1970s with the advent of AOD (argon–oxygen decarburization) melting practice that
reduced the cost differential between standard (not low carbon) and L grades. These
low-carbon grades (304L, 316L) have been widely used in applications where inter-
granular attack and intergranular stress corrosion cracking are a concern. These
forms of corrosion attack are discussed later in the chapter.
Austenitic stainless steels are used in a wide range of applications, including struc-
tural support and containment, architectural uses, kitchen equipment, and medical
products. They are widely used not only because of their corrosion resistance but
because they are readily formable, fabricable, and durable. Some highly alloyed
grades are used for very high temperature service [above 1000°C (1830°F)] for appli-
cations such as heat-treating baskets. In addition to higher chromium levels, these
alloys normally contain higher levels of silicon (and sometimes aluminum) and car-
bon, to maintain oxidation and/or carburization resistance and strength, respectively.
It should be pointed out that the common austenitic stainless steels are not an
appropriate choice in some common environments such as seawater or other
 STANDARD ALLOYS AND CONSUMABLES 143

chloride-containing media, or in highly caustic environments. This is due to their

susceptibility to stress corrosion cracking, a phenomenon that afflicts the base metal,
HAZ, and weld metal in these alloys. Care should be taken when selecting stainless
steels that will be under significant stress in these environments. The aspects and
consequences of stress corrosion cracking will be described later.

6.1 STANDARD ALLOYS AND CONSUMABLES

Austenitic stainless steels include both the 200 series and 300 series alloys, as des-
ignated by the American Iron and Steel Institute (AISI). The 200 series alloys con-
tain high levels of carbon, manganese, and nitrogen and are used in specialty
applications, such as where galling resistance is required. These alloys also have
lower nickel content than the 300 series alloys to balance the high carbon and nitro-
gen levels. The 300 series alloys are by far the most widely used of the austenitic
grades. A list of the most common 300 series alloys is provided in Table 6.1. A more
comprehensive list of alloys is provided in Appendix 1.
The most widely used alloys, Types 304, 316, 321, and 347 and their variants, are
of the “18–8” type with nominal values of 18Cr and 8–10Ni. L grades represent low-
carbon variants with a nominal carbon level of 0.03 wt%. These alloys have
improved resistance to intergranular attack in corrosive environments. H grades have
carbon levels approaching 0.1 wt%. These alloys are used at elevated temperature
since they have higher elevated temperature strength than standard or L grades. The
N grades have nitrogen added as an intentional addition to levels as high as 0.20 wt %
in the 300 series alloys (304N, 316N) and to even higher levels in alloys that contain
high manganese contents. (Manganese increases the solubility of nitrogen in the
austenite phase.) The higher nitrogen improves the strength, galling resistance, and
pitting corrosion resistance of austenitic stainless steels.
Alloys containing titanium and niobium, such as Types 321 and 347, are known
as stabilized grades since the addition of these two elements stabilizes the alloy
against the formation of M23C6 chromium carbides. Since niobium and titanium both
form stable MC-type carbides at elevated temperature, the formation of chromium-
rich carbides is restricted. The addition of these elements at levels up to 1.0 wt%
effectively reduces the matrix carbon content and makes chromium carbide precipi-
tation more difficult. This reduces the possibility for sensitization that can lead to
intergranular corrosion in austenitic alloys. This phenomenon is described in more
detail in Section 6.6.
Many other specialty alloys exist, including the superaustenitic grades. These
alloys are described separately in a later section because of their unique characteris-
tics and weldability concerns.
Austenitic stainless steel filler metals are listed in Table 6.2. Note that this table
is divided into three parts, reflecting three AWS specifications for consumables
based on type: (1) AWSA5.4 for covered electrodes (SMAW), (2) AWS A5.9 for bare
wire and tubular metal-cored electrodes (GTAW and GMAW), and (3) AWS A5.22
for gas-shielded flux-cored electrodes (FCAW). Many of the same consumables
144

TABLE 6.1 Composition of Standard Wrought Austenitic Stainless Steels

Composition (wt%)a
Type UNS No. C Mn P S Si Cr Ni Mo N Other
201 S20100 0.15 5.5–7.5 0.06 0.03 1.0 16.0–18.0 3.5–5.5 — 0.25 —
302 S30200 0.15 2.0 0.045 0.03 1.0 17.0–19.0 8.0–10.0 — — —
304 S30400 0.08 2.0 0.045 0.03 1.0 18.0–20.0 8.0–10.5 — — —
304L S30403 0.03 2.0 0.045 0.03 1.0 18.0–20.0 8.0–12.0 — — —
304H S30409 0.04–0.1 2.0 0.045 0.03 1.0 18.0–20.0 8.0–10.5 — — —
308 S30800 0.08 2.0 0.045 0.03 1.0 19.0–21.0 10.0–12.0 — — —
309 S30900 0.20 2.0 0.045 0.03 1.0 22.0–24.0 12.0–15.0 — — —
310 S31000 0.25 2.0 0.045 0.03 1.0 24.0–26.0 19.0–22.0 — — —
316 S31600 0.08 2.0 0.045 0.03 1.0 16.0–18.0 10.0–14.0 2.0–3.0 — —
316L S31603 0.03 2.0 0.045 0.03 1.0 16.0–18.0 10.0–14.0 2.0–3.0 — —
317 S31700 0.08 2.0 0.045 0.03 1.0 18.0–20.0 11.0–15.0 3.0–4.0 — —
321 S32100 0.08 2.0 0.045 0.03 1.0 17.0–19.0 9.0–12.0 — — Ti: 5  C-0.70
330 S33000 0.10 2.0 0.045 0.03 0.75–1.5 17.0–20.0 34.0–37.0 — — —
347 S34700 0.08 2.0 0.045 0.03 1.0 17.0–19.0 9.0–13.0 — — Nb: 10  C-1.00
a
A single value is a maximum.
 TABLE 6.2 Austenitic Stainless Steel AWS Filler Metal Classifications
Composition (wt%)a
Type UNS No. C Mn P S Si Cr Ni Mo N Other
Part 1: Covered Electrodes from AWS A5.4
219 W32310 0.06 8.0–10.0 0.04 0.03 1.00 19.0–21.5 5.5–7.0 0.75 0.10–0.30 —
308 W30810 0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.75 — —
308H W30810 0.04–0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.75 — —
308L W30813 0.04 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.75 — —
309 W30910 0.15 0.5–2.5 0.04 0.03 1.0 22.0–25.0 12.0–14.0 0.75 — —
309L W30917 0.04 0.5–2.5 0.04 0.03 1.0 22.0–25.0 12.0–14.0 0.75 — —
310 W31010 0.08–0.20 1.0–2.5 0.03 0.03 0.75 25.0–28.0 20.0–22.5 0.75 — —
316 W31610 0.08 0.5–2.5 0.04 0.03 1.0 17.0–20.0 11.0–14.0 2.0–3.0 — —
316H W31610 0.04–0.08 0.5–2.5 0.04 0.03 1.0 17.0–20.0 11.0–14.0 2.0–3.0 — —
316L W31613 0.04 0.5–2.5 0.04 0.03 1.0 17.0–20.0 11.0–14.0 2.0–3.0 — —
317 W31710 0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 12.0–14.0 3.0–4.0 — —
317L W31713 0.04 0.5–2.5 0.04 0.03 1.0 18.0–21.0 12.0–14.0 3.0–4.0 — —
330 W88331 0.18–0.25 1.0–2.5 0.04 0.03 0.90 14.0–17.0 33.0–37.0 0.75 — —
347 W34710 0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.75 — Nb: 8  C-1.00

Part 2: Bare Electrodes, Bare Rods, Tubular Metal-Cored Electrodes, and Strips from AWS A5.9
219 S21980 0.05 8.0–10.0 0.03 0.03 1.00 19.0–21.5 5.5–7.0 0.75 0.10–0.30 —
308 S30880 0.08 1.0–2.5 0.03 0.03 0.30–0.65 19.5–22.0 9.0–11.0 0.75 — —
308H S30880 0.04–0.08 1.0–2.5 0.03 0.03 0.30–0.65 19.5–22.0 9.0–11.0 0.75 — —
308L S30883 0.03 1.0–2.5 0.03 0.03 0.30–0.65 19.5–22.0 9.0–11.0 0.75 — —
308Si S30881 0.08 1.0–2.5 0.03 0.03 0.65–1.00 19.5–22.0 9.0–11.0 0.75 — —
308LSi S30888 0.03 1.0–2.5 0.03 0.03 0.65–1.00 19.5–22.0 9.0–11.0 0.75 — —
309 S30980 0.12 1.0–2.5 0.03 0.03 0.30–0.65 23.0–25.0 12.0–14.0 0.75 — —
145

309L S30983 0.03 1.0–2.5 0.03 0.03 0.30–0.65 23.0–25.0 12.0–14.0 0.75 — —
(continued)
 TABLE 6.2 (Continued)
146
Composition (wt%)a
Type UNS No. C Mn P S Si Cr Ni Mo N Other
309Si S30981 0.12 1.0–2.5 0.03 0.03 0.65–1.00 23.0–25.0 12.0–14.0 0.75 — —
309LSi S30988 0.03 1.0–2.5 0.03 0.03 0.65–1.00 23.0–25.0 12.0–14.0 0.75 — —
310 S31080 0.08–0.15 1.0–2.5 0.03 0.03 0.30–0.65 25.0–28.0 20.0–22.5 0.75 — —
316 S31680 0.08 1.0–2.5 0.03 0.03 0.30–0.65 18.0–20.0 11.0–14.0 2.0–3.0 — —
316H S31680 0.04–0.08 1.0–2.5 0.03 0.03 0.30–0.65 18.0–20.0 11.0–14.0 2.0–3.0 — —
316L S31683 0.03 1.0–2.5 0.03 0.03 0.30–0.65 18.0–20.0 11.0–14.0 2.0–3.0 — —
316Si S31681 0.08 1.0–2.5 0.03 0.03 0.65–1.00 18.0–20.0 11.0–14.0 2.0–3.0 — —
316LSi S31688 0.03 1.0–2.5 0.03 0.03 0.65–1.00 18.0–20.0 11.0–14.0 2.0–3.0 — —
317 S31780 0.08 1.0–2.5 0.03 0.03 0.30–0.65 18.5–20.5 13.0–15.0 3.0–4.0 — —
317L S31783 0.03 1.0–2.5 0.03 0.03 0.30–0.65 18.5–20.5 13.0–15.0 3.0–4.0 — —
330 N08331 0.18–0.25 1.0–2.5 0.03 0.03 0.30–0.65 15.0–17.0 34.0–37.0 0.75 — —
347 S34780 0.08 1.0–2.5 0.03 0.03 0.30–0.65 19.0–21.5 9.0–11.0 0.75 — Nb: 10  C-1.0
347Si S34788 0.08 1.0–2.5 0.03 0.03 0.65–1.00 19.0–21.5 9.0–11.0 0.75 — Nb: 10  C-1.0

Part 3: Gas-Shielded Flux-Cored Electrodes from AWS A5.22b

308 W30831 0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.5 — —
308L W30835 0.04 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.5 — —
308H W30831 0.04–0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.5 — —
309 W30931 0.10 0.5–2.5 0.04 0.03 1.0 22.0–25.0 12.0–14.0 0.5 — —
309L W30935 0.04 0.5–2.5 0.04 0.03 1.0 22.0–25.0 12.0–14.0 0.5 — —
310 W31031 0.20 1.0–2.5 0.03 0.03 1.0 25.0–28.0 20.0–22.5 0.5 — —
316 W31631 0.08 0.5–2.5 0.04 0.03 1.0 17.0–20.0 11.0–14.0 2.0–3.0
316L W31635 0.04 0.5–2.5 0.04 0.03 1.0 17.0–20.0 11.0–14.0 2.0–3.0 — —
317L W31735 0.04 0.5–2.5 0.04 0.03 1.0 18.0–21.0 12.0–14.0 3.0–4.0 — —
347 W34731 0.08 0.5–2.5 0.04 0.03 1.0 18.0–21.0 9.0–11.0 0.5 — Nb: 8  C-1.0
a
A single value is a maximum.
b
Self-shielded flux-cored stainless steel electrodes are also included in the AWS A5.22 standard. They have virtually identical composition limits to those of the gas-shielded
electrodes except that the Cr limits tend to be slightly higher in the ferrite-containing grades, to compensate for expected higher nitrogen in the self-shielded deposit.
 PHYSICAL AND MECHANICAL METALLURGY 147

appear in all three specifications, but the composition limits may be slightly different
based on consumable type.

6.2 PHYSICAL AND MECHANICAL METALLURGY

Austenitic stainless steels are formulated and thermo-mechanically processed such

that the microstructure is primarily austenite. Depending on the balance of ferrite-
promoting to austenite-promoting elements, the wrought or cast microstructure will
be either fully austenitic or a mixture of austenite and ferrite. Two examples of
wrought austenitic stainless steel microstructures are shown in Figure 6.1. In one
case, the microstructure consists of equiaxed austenite grains. In the other case,
some residual high-temperature ferrite (delta ferrite) is aligned along the rolling
direction. This ferrite results from the segregation of ferrite-promoting elements (pri-
marily chromium) during solidification and thermo-mechanical processing. It is usu-
ally present in relatively low volume fraction (less than 2 to 3%). Although not
considered deleterious in most applications, the presence of ferrite in the wrought
microstructure can reduce the ductility and, potentially, the toughness of austenitic
stainless steels. It can also be a preferential site for the precipitation of M23C6 car-
bides and sigma phase, of which the latter is an embrittling agent in stainless steels.
The transformation behavior of austenitic stainless steels can be described using the
Fe–Cr–Ni pseudobinary diagram at 70% constant iron [1], as described in Chapter 2.
This diagram is reproduced here as Figure 6.2. Note that primary solidification of
austenitic stainless steels can occur as either austenite or ferrite. The demarcation
between these two primary phases of solidification is at approximately 18Cr–12Ni in
the ternary system. At higher chromium/nickel ratios, primary solidification occurs as
delta ferrite and at lower ratios as austenite. Note that there is a small triangular region
within the solidification temperature range where austenite, ferrite, and liquid coexist.
Alloys that solidify as austenite to the left of this triangular region are stable as austen-
ite upon cooling to room temperature. However, when alloys solidify as ferrite, they
may be either fully ferritic or consist of a mixture of ferrite and austenite at the end of
solidification. Because of the slope of the ferrite and austenite solvus lines, most or all
of the ferrite transforms to austenite under equilibrium cooling conditions, as can be
seen for a nominal 20Cr–10Ni alloy where the structure becomes fully austenitic upon
cooling to 1000°C (1830°F). For the rapid cooling conditions experienced during
welding, this transformation is suppressed and some ferrite will remain in the
microstructure. The implications of these are described in more detail in Section 6.3.1.
The effect of other alloying elements on the phase equilibrium in austenitic stain-
less steels can be determined using phase diagrams generated using computational
thermodynamics, such as ThermCalc™ [2]. As an example, a Fe–18Cr–10Ni–1.5Mn–
0.5Si–0.04N system with variable carbon and a Fe–10Ni–1.5Mn–0.5Si 0.04C–0.04N
system with variable chromium are shown in Figure 6.3. These diagrams are very sim-
ilar to Figure 2.5a, but additionally, show the carbide and nitride precipitation regimes.
A variety of precipitates may be present in austenitic stainless steels, depending
on composition and heat treatment. A list of those precipitates, their structure, and
148 AUSTENITIC STAINLESS STEELS

Figure 6.1 Microstructure of Type 304 plate: (a) fully austenitic; (b) austenite with ferrite
stringers.

stoichiometry is provided in Table 6.3. Carbides are present in virtually every

austenitic stainless steel, since chromium is a strong carbide former. Additions of
other carbide formers, including Mo, Nb, and Ti, also promote carbide formation.
The nature of carbide formation, including the effect of composition and the tem-
perature range of formation, is quite complex. The reader is referred to other, more
comprehensive sources for more details on carbide formation [3,4].
 PHYSICAL AND MECHANICAL METALLURGY 149

Nickel content (wt%)

25 20 15 10 5 0
1600

1400 2600
+L
+L
+
+L +



Temperature (°C)

Temperature (°F)
1200 2200
 solvus
1000 1800

solvus
+
+
800 1400
+

600 1000

+

400
5 10 15 20 25 30
Chromium content (wt%)

Figure 6.2 Pseudobinary section of the Fe–Cr–Ni system at 70% iron. (From Lippold and
Savage [1]. Courtesy of the American Welding Society.)

Precipitation of M23C6 carbides has received considerable attention because of

its effect on corrosion resistance. As shown in Figure 6.4, these carbides precipitate
very rapidly along grain boundaries in the temperature range 700 to 900°C (1290
to 1650°F). At slightly longer times, the presence of these grain boundary carbides
can lead to intergranular corrosion when exposed to certain environments [5]. This
precipitation reaction is accelerated in alloys that are strengthened by cold work.
Sigma, chi, eta, G, and Laves phases (Table 6.3) may also form in austenitic stain-
less steels, particularly in those that contain additions of Mo, Nb, and Ti. Typically,
these phases form after longtime exposure at elevated temperature and result in
embrittlement of the steel. An example of the effect of sigma phase on the impact
toughness of Fe–Cr–Ni alloys is shown in Figure 6.5. Note that less than 5% sigma
is effective in reducing the impact toughness by over 50%.

6.2.1 Mechanical Properties

Minimum room-temperature mechanical properties of a number of austenitic stain-
less steels are provided in Table 6.4. These properties reflect the hot-finished and
annealed condition. Considerable strength can be imparted to these alloys by cold
working. In general, the austenitic stainless steels cannot be substantially strength-
ened by either precipitation or transformation. Some specialty alloys containing
higher nickel contents and titanium additions can be strengthened by precipitation of
the gamma-prime precipitate, Ni3(Al,Ti), that is a common strengthening agent in
nickel-base superalloys. These alloys are discussed in Chapter 8. It is also possible
to form martensite in some austenitic grades, but this occurs under only very special
150 AUSTENITIC STAINLESS STEELS

1600
Liquid +
Liquid Liquid + 

1400

+

Liquid +
+ 
1200

Temperature (°C)

1000

 + M23C6 + Cr2N
800  + M23C6

 + Cr2N + M23C6 + 
600
 + M23C6 + Cr2N + 

400
0 0.02 0.04 0.06 0.08 0.10
(a) C (wt %)

1600
Liquid +  Liquid Liquid +

1400
Liquid +
+ 

1200
+
Temperature (°C)

+  + M23C6
1000  + M23C6 + Cr2N

 + M23C6
800  + M23C6 + 

 + M23C6 +  + Cr2N
600
 + M23C6 +  + Cr2N + 

400
15 16 17 18 19 20 21 22 23 24 25
(b) Cr (wt %)

Figure 6.3 Phase diagrams calculated using ThermoCalc: (a) Fe–18Cr–10Ni–1.5Mn–

0.5Si–0.04N, variable C; (b) Fe–10Ni–1.5Mn–0.5Si–0.04C–0.04N, variable Cr. (Courtesy of
Antonio Ramirez.)
 WELDING METALLURGY 151

TABLE 6.3 Precipitates in Austenitic Stainless Steels

Precipitate Crystal Structure Lattice Parameters (nm) Stoichiometry
MC FCC a  0.424–0.447 TiC, NbC
M6C Diamond cubic a  1.062–1.128 (FeCr)3Mo3C, Fe3Nb3C,
Mo5SiC
M23C6 FCC a  1.057–1.068 (Cr,Fe)23C6,
(Cr,Fe,Mo)23C6
NbN FCC a  0.440 NbN
Z phase Tetragonal a  0.307, c  0.7391 CrNbN
Sigma phase Tetragonal a  0.880, c  0.454 Fe–Ni–Cr–Mo
Laves phase (η) Hexagonal a  0.473, c  0.772 Fe2Mo, Fe2Nb
Chi phase (χ) BCC a  0.8807–0.8878 Fe36Cr12Mo10
G phase FCC a  1.12 Ni16Nb6Si7, Ni16Ti6Si7
R Hexagonal a  1.0903, c  1.9342 Mo–Co–Cr
Rhombohedral a  0.9011 α  74°27.5 Mo–Co–Cr
ε Nitride (Cr2N) Hexagonal a  0.480, c  0.447 Cr2N
Ni3Ti Hexagonal a  0.9654, c  1.5683 Ni3Ti
Ni3(Al,Ti) FCC a  0.681 Ni3Al

1100 2012
1000 1832
Precipitation temperature (°C)

900 Initiation of 1652

grain-boundary
Temperature (°F)

800 precipitation 1472

700 Onset of 1292

intergranular
corrosion 1112
600

500 932

400 752
0.015 0.15 1.5 15 150 1500
Time (h)

Figure 6.4 Precipitation of M23C6 carbides in Type 304 stainless steel with 0.05 wt% car-
bon. (From Cihal [5].)

conditions. Martensite has been observed in very heavily cold-worked alloys and/or
when materials are cooled to cryogenic temperatures [3,7].

6.3 WELDING METALLURGY

The room-temperature microstructure of the fusion zone of austenitic stainless

steels is dependent both on the solidification behavior and subsequent solid-state
152 AUSTENITIC STAINLESS STEELS

Charpy imapct (keyhole notch) (ft-lb.)

50 67.8
%Ni %Cr Annealed Cold worked
40 35 30 x 54.2
35 25
30 25
30 25 30 40.7

Joules
25 25
20 25
20 27.1

10 x 13.6
x

0
4 8 12 16 20 24 28
Sigma (%)

Figure 6.5 Effect of sigma phase on the room-temperature impact strength of Fe–Cr–Ni
alloys. (From Talbot and Furman [6]. Courtesy of ASM International.)

Table 6.4 Minimum Room-Temperature Mechanical Properties of Wrought

Austenitic Stainless Steels
Tensile Strength Yield Strength
Elongation Reduction
Alloy MPa ksi MPa ksi (%) in Area (%)
302 515 75 205 30 40 50
304 515 75 205 30 40 50
304L 480 70 170 25 40 50
308 515 75 205 30 40 50
309 515 75 205 30 40 50
310 515 75 205 30 40 50
316 515 75 205 30 40 50
316L 480 70 170 25 40 50
317 515 75 205 30 40 50
321 515 75 205 30 40 50
330 480 70 205 30 30 —
347 515 75 205 30 40 50
Source: ASM Handbook, Vol. 6, p. 468.

transformations. All stainless steels solidify with either ferrite or austenite as the pri-
mary phase. The austenitic stainless steels may solidify as primary ferrite or primary
austenite, depending on the specific composition. Small changes in composition
within a given alloy system may promote a shift from primary ferrite to primary
austenite. The composition range of many austenitic stainless steels is broad enough
that both solidification modes are possible. Following solidification, additional trans-
formations can occur in the solid state upon cooling to room temperature. These
transformations are most important in alloys undergoing primary ferrite solid-
ification, since most of the ferrite will transform to austenite.
 WELDING METALLURGY 153

6.3.1 Fusion Zone Microstructure Evolution

There are four solidification and solid-state transformation possibilities for austenitic
stainless steel weld metals. These reactions are listed in Table 6.5 and related to the
Fe–Cr–Ni phase diagram in Figure 6.6. Note that A and AF solidification modes are
associated with primary austenite solidification, whereby austenite is the first phase
to form upon solidification. Types FA and F solidification have delta ferrite as the
primary phase. Following solidification, additional microstructural modification

TABLE 6.5 Solidification Types, Reactions, and Resultant Microstructures

Solidification
Type Reaction Microstructure
A L→L
A→A Fully austenitic, well-defined solidification
structure
AF L→L
A→L
A
Ferrite at cell and dendrite boundaries
(A
F)eut → A
Feut
FA L→L
F→L
F
Skeletal and/or lathy ferrite resulting from
(F
A)per/eut → F
A ferrite-to-austenite transformation
F L→L
F→ Acicular ferrite or ferrite matrix with grain
F→F
A boundary austenite and Widmanstätten
side plates

Creq /Nieq

A AF FA F

Liquid

L+

L+
Ferrite
L++
+

Temperature

solvus



Fully
austenitic

Austenite +
eutectic ferrite

Austenite + Austenite + Austenite + Ferrite +

skeletal lathy acicular Widmanstätten
ferrite ferrite ferrite austenite

Figure 6.6 Relationship of solidification type to the pseudobinary phase diagram.

154 AUSTENITIC STAINLESS STEELS

occurs in the solid state for the FA and F types, due to the instability of the ferrite at
lower temperatures. The various microstructures that are possible in austenitic stain-
less steel weld metals and their evolution are described in the following sections.

6.3.1.1 Type A: Fully Austenitic Solidification When solidification occurs

as primary austenite, two weld metal microstructures are possible. If the microstruc-
ture is fully austenitic at the end of solidification, it will remain austenitic upon cool-
ing to room temperature and exhibit a distinct solidification structure when viewed
metallographically. This is defined as Type A solidification and is shown schemati-
cally in Figure 6.7. An example of Type A solidification is shown metallographically
in Figure 6.8. Note that the solidification substructure (cells and dendrites) is read-
ily apparent in this microstructure. This is characteristic of solidification as primary
austenite due to the segregation of alloying and impurity elements that occurs dur-
ing solidification and the relatively low diffusivity of these elements at elevated tem-
perature, which preserves the segregation profile that develops during solidification.
When alloys such as Types 304 and 316 solidify as Type A, Cr and Mo have been
shown to partition to the cell and dendrite boundaries.

Solidification
Liquid
grain boundary

    
 

Liquid droplet
Migrated grain
Austenite
boundary

Figure 6.7 Type A solidification, fully austenitic. (From Katayama et al. [14]. Courtesy of
the Japanese Welding Research Institute.)
 WELDING METALLURGY 155

Figure 6.8 Fusion zone microstructure resulting from fully austenitic (Type A) solid-
ification.

6.3.1.2 Type AF Solidification If some ferrite forms at the end of the primary
austenite solidification process via a eutectic reaction, solidification is termed Type
AF. This occurs if sufficient ferrite-promoting elements (primarily Cr and Mo) par-
tition to the solidification subgrain boundaries during solidification to promote the
formation of ferrite as a terminal solidification product. This is thought to occur by
a eutectic reaction and is represented by the three-phase triangular region of the
phase diagram in Figures 6.2 and 6.6. The ferrite that forms along the boundary is
relatively stable and resists transformation to austenite during weld cooling since it
is already enriched in ferrite-promoting elements. A schematic of AF solidification
is shown in Figure 6.9. An example of a microstructure that exhibits ferrite along
solidification subgrain boundaries is shown in Figure 6.10. Note that because this is
primary austenite solidification, the solidification substructure is readily apparent.

6.3.1.3 Type FA Solidification When solidification occurs as primary ferrite,

there are also two possibilities. If some austenite forms at the end of solidification, it
is termed Type FA. This austenite forms via a peritectic–eutectic reaction and exists at
the ferrite solidification boundaries at the end of solidification. This reaction has been
studied extensively by David et al. [8,9], Lippold and Savage [1,10,11], Brooks et al.
[12], Arata et al. [13], Katayama et al. [14], Leone and Kerr [15], and others [16,17].
Based on these investigations, the following solidification and transformation
sequence occurs to give rise to the ferrite morphologies resulting from FA
solidification (Figures 6.11 and 6.12).
156 AUSTENITIC STAINLESS STEELS

  

Liquid droplet

Austenite

-ferrite
Migrated grain
boundary

Figure 6.9 Fusion zone microstructure resulting from Type AF solidification. (From
Katayama et al. [14]. Courtesy of the Japanese Welding Research Institute.)

1. At the end of primary ferrite solidification, a peritectic–eutectic reaction results

in the formation of austenite along the ferrite cell and dendrite boundaries. This
reaction occurs within and along the triangular three-phase region shown in
Figures 6.2 and 6.6. It is called a peritectic–eutectic reaction because it is com-
position dependent and results from a transition from a peritectic reaction in the
Fe–Ni system to a eutectic reaction in the Fe–Cr–Ni system (see Figure 2.4).
2. When solidification is complete, the microstructure consists of primary ferrite
dendrites with an interdendritic layer of austenite. The amount of austenite
that is present depends on the solidification conditions and the Creq/Nieq value.
As Creq/Nieq increases, the amount of austenite decreases until solidification is
entirely ferritic. At this point the solidification type shifts from FA to F.
3. As the weld metal cools through the two-phase delta ferrite
austenite field,
the ferrite becomes increasingly unstable and the austenite begins to consume
the ferrite via a diffusion-controlled reaction. There was some early debate
about the nature of this reaction [1,8,9,11,15], but it is now generally agreed
 WELDING METALLURGY 157

Figure 6.10 Fusion zone microstructure resulting from Type AF solidification.

that diffusion across the austenite–ferrite interface controls the rate and nature
of the transformation.
4. When weld cooling rates are moderate and/or when the Creq/Nieq is low but
still within the FA range (Figure 6.6), a vermicular, or skeletal, ferrite mor-
phology results. This is a consequence of the advance of the austenite con-
suming the ferrite until the ferrite is sufficiently enriched in ferrite-promoting
elements (chromium and molybdenum) and depleted in austenite-promoting
elements (nickel, carbon, and nitrogen) that it is stable at lower temperatures
where diffusion is limited. This skeletal microstructure is shown schematically
in Figure 6.11a and in the micrograph in Figure 6.12a.
5. When cooling rates are high and/or when the Creq/Nieq increases within the FA
range in Figure 6.6, a lathy ferrite morphology results. The lathy morphology
forms in place of the skeletal morphology due to restricted diffusion during the
ferrite–austenite transformation. When diffusion distances are reduced it is
more efficient for the transformation to proceed as more tightly spaced laths,
resulting in a residual ferrite pattern that cuts across the original dendrite or
cell growth direction. This is shown schematically in Figure 6.11b and in the
micrograph in Figure 6.12b.
6. When solidification and cooling rates are extremely high, such as during laser
or electron beam welding, a complete transformation from ferrite to austenite
may be possible due to a diffusionless, “massive” transformation. A shift in pri-
mary solidification mode from ferrite to austenite may also occur at high
solidification rates. This behavior is described in more detail in Section 6.5.1.5.
158 AUSTENITIC STAINLESS STEELS

Liquid Solidification
grain boundary

-ferrite

Austenite
Austenite

-ferrite

Migrated grain
boundary Migrated grain
boundary

(a) (b)

Figure 6.11 Type FA solidification: (a) skeletal ferrite; (b) lathy morphology. (From
Katayama et al. [14]. Courtesy of the Japanese Welding Research Institute.)

Considerable research into the ferrite–austenite transformation sequence was

conducted in the late 1970s and 1980s. For more details regarding this transforma-
tion, the reader is referred to the review by Brooks and Thompson [18].

6.3.1.4 Type F Solidification If solidification occurs completely as ferrite, it

is termed Type F. In this case the microstructure is fully ferritic at the end of solid-
ification, as shown in Figure 6.6. When the weld metal cools below the ferrite solvus,
austenite forms within the microstructure, usually first at the ferrite grain boundaries.
Because the structure was fully ferritic in the solid state between the solidus and fer-
rite solvus, diffusion eliminates most or all composition gradients resulting from
solidification, and thus, when the transformation starts, the microstructure consists
 WELDING METALLURGY 159

Figure 6.12 Fusion zone microstructure resulting from FA solidification: (a) skeletal ferrite
morphology; (b) lathy ferrite morphology.

of large relatively homogeneous ferrite grains. The degree of transformation to

austenite again depends on Creq/Nieq and cooling rate. At low Creq/Nieq values within
the F range (Figure 6.6), the transformation begins at a higher temperature, and at
low to moderate weld cooling rates, much of the ferrite is consumed. With higher
cooling rates, diffusion is suppressed and the austenite will not consume as much of
the ferrite. Similarly, if the Creq/Nieq value is increased within the F range, the ferrite
160 AUSTENITIC STAINLESS STEELS

solvus is depressed and transformation will occur at lower temperatures. In both

cases, weld metals with high ferrite contents will result.
The microstructure that forms as a result of Type F solidification in austenitic
stainless steels again is a function of composition and cooling rate. At low Creq/Nieq
values within the F range (Figure 6.6) an acicular ferrite structure will form within
the ferrite grains. This structure is shown in the schematic in Figure 6.13a. Note that
continuous austenite networks are present at the prior ferrite grain boundaries and
that the acicular ferrite is no longer contained within the bounds of the original fer-
rite dendrites, as during FA solidification with lathy ferrite formation (Figure 6.11b).
This occurs because of the absence of austenite within the ferrite grains during Type
F solidification. The structure is completely ferritic in the solid state before the trans-
formation to austenite begins. When this structure cools below the ferrite solvus,
austenite first forms at the ferrite grain boundary, but the transformation front breaks
down and parallel needles of austenite form within the ferrite. As in the case of FA

Liquid droplet Liquid droplet

-ferrite
Migrated
grain boundary

-ferrite

-ferrite
Austenite Austenite

(a) (b)

Figure 6.13 Schematic of Type F solidification: (a) acicular ferrite; (b) ferrite and Widman-
stätten austenite. (From Katayama et al. [14]. Courtesy of the Japanese Welding Research
Institute.)
 WELDING METALLURGY 161

solidification with lathy ferrite, the restriction of long-range diffusion at the lower
transformation temperature forces the transformation to occur over shorter distances.
This produces the acicular structure shown in Figure 6.13a.
At higher Creq/Nieq values (given the same cooling rate) the microstructure will
consist of a ferrite matrix with grain boundary austenite and Widmanstätten austen-
ite plates that nucleate at the grain boundary austenite or within the ferrite grains.
This microstructure is shown schematically in Figure 6.13b and in a micrograph in
Figure 6.14. In this case, transformation does not occur entirely across the ferrite
grain. Initial austenite again forms at the ferrite grain boundary, but transformation
across the entire grain is suppressed by lower diffusion rates and lower driving force
(the equilibrium microstructure contains more ferrite). This again can be understood
from the pseudobinary diagram in Figure 6.6. As Creq/Nieq increases, the ferrite
solvus decreases and the equilibrium ferrite content increases, thus reducing the
driving force for the ferrite-to-austenite transformation and the temperature at which
the transformation begins.
In practice, Type F solidification is very unusual in austenitic stainless steel weld
metals. Most filler metals are formulated such that solidification occurs in the FA
mode, with weld metal ferrite contents ranging from 5 to 20 FN (Ferrite Number).
Only highly alloyed filler metals such as Type 309LMo and Type 312* (30Cr–10Ni)

Figure 6.14 Fusion zone microstructure resulting from F solidification: Widmanstätten

austenite nucleates from austenite along ferrite grain boundaries. This microstructure is very
unusual in austenitic stainless steels.

*Type 312 can be considered as an austenitic or as a duplex filler metal. The microstructure of Type 312
deposits is more representative of duplex stainless steel weld metals (30 to 80 FN) than of austenitic stain-
less steel weld metals.
162 AUSTENITIC STAINLESS STEELS

would be expected to exhibit microstructures with higher levels of ferrite. Type F

solidification (such as that shown in Figure 6.14) is more characteristic of the duplex
stainless steels, as described in Chapter 7.

6.3.2 Interfaces in Single-Phase Austenitic Weld Metal

It is important to understand the nature of the various boundaries, or interfaces, that
are present in austenitic stainless steel weld metal, since many of the defects associ-
ated with the fusion zone both during fabrication and in service are associated with
these boundaries. Boundaries are especially evident in weld metals that solidify in
the A or AF mode, since the solidification structure is clearly apparent after polish-
ing and etching. At least three different boundary types can be observed metallo-
graphically [19]. These are shown schematically in Figure 6.15 and described in the
following sections.

6.3.2.1 Solidification Subgrain Boundaries The solidification subgrains

represent the finest structure that can be resolved in the optical microscope. These
subgrains are normally present as cells or dendrites and the boundary separating
adjacent subgrains is known as a solidification subgrain boundary (SSGB). These
boundaries are evident in the microstructure because their composition is different

Fusion zone

Solidification
subgrain boundary

Migrated grain
boundary
Solidification grain
boundary

Figure 6.15 Schematic of boundaries observed in weld metals that solidify as primary
austenite (A and AF mode).
 Next Page
WELDING METALLURGY 163

from that of the bulk microstructure. Solute redistribution that creates this composi-
tional gradient at the SSGB is dictated by Case 2 (microscopic) solute redistribution,
also referred to as Scheil partitioning.
There is little crystallographic misorientation across the SSGB, and these bound-
aries are characterized crystallographically as low-angle boundaries. The low angu-
lar misorientation (approaching zero) results from the fact that subgrain growth
during solidification occurs along preferred crystallographic directions (or easy
growth directions). In FCC and BCC metals, these are 〈100〉 directions. Because of
this, the dislocation density along the SSGB is generally low since there is not a
large structural misorientation to accommodate.

6.3.2.2 Solidification Grain Boundaries The solidification grain boundary

(SGB) results from the intersection of packets, or groups, of subgrains. Thus, SGBs
are the direct result of competitive growth that occurs during solidification along the
trailing edge of the weld pool. Because each of these packets of subgrains has a
different growth direction and orientation, their intersection results in a boundary with
high angular misorientation. These are often called high-angle grain boundaries. This
misorientation results in the development of a dislocation network along the SGB.
The SGB also exhibits a compositional component resulting from solute redistri-
bution during solidification. This redistribution can be modeled using Case 3 (macro-
scopic) solidification boundary conditions and often results in high concentrations of
solute and impurity elements at the SGBs. These compositions may lead to the for-
mation of low-melting liquid films along the SGBs at the conclusion of solidification
that can promote weld solidification cracking. When weld solidification cracking
occurs in stainless steels, it is almost always along SGBs.

6.3.2.3 Migrated Grain Boundaries The SGB that forms at the end of
solidification has both a compositional and a crystallographic component. In some sit-
uations, it is possible for the crystallographic component of the SGB to migrate away
from the compositional component. This new boundary that carries with it the high-
angle misorientation of the “parent” SGB is called a migrated grain boundary (MGB).
The driving force for migration is the same as for simple grain growth in base met-
als, a lowering of boundary energy. The original SGB is quite tortuous since it forms
from the intersection of opposing cells and dendrites. The crystallographic boundary
can lower its energy by straightening and, in the process, it pulls away from the orig-
inal SGB. Further migration of the boundary is possible during reheating, such as dur-
ing multipass welding. Because it carries the crystallographic misorientation of the
SGB with it, the MGB represents a high-angle boundary, normally with misorienta-
tions greater than 30°. The composition of the boundary varies locally, depending on
the composition of the microstructure where it has migrated. It is also possible that
some segregation can occur along MGBs, possibly by a “sweeping” mechanism.
MGBs are most prevalent in fully austenitic weld metals. When the weld metal
undergoes AF solidification, ferrite forms at the end of solidification along the SSGBs
and SGBs. This ferrite is effective in “pinning” the crystallographic component of the
SGB, thus preventing it from migrating away from the parent SGB. In this case, an