WSRC Ms 2001 00544
WSRC Ms 2001 00544
net/publication/225107859
Considerations for the weldability of types 304L and 316L stainless steel
CITATIONS READS
34 350
2 authors, including:
Paul Korinko
Savannah River National Laboratory
86 PUBLICATIONS 169 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Paul Korinko on 06 May 2020.
P. S. Korinko
S. H. Malene
Summary
of penetration, two distinct weld characteristics related to the chemical composition of the base
material is reviewed. The propensity for cracking is determined primarily by the solidification
mode and the amount of residual tramp elements such as phosphorous and sulfur. High sulfur
levels can lead to weld centerline cracking and Heat Affected Zone (HAZ) cracking while very
low sulfur levels (less than ~50 ppm) in types 304L and 316L are associated with lack of
penetration weld defects and a distinct loss in puddle control during fusion welding. A
calculated Creqivalence to Nieq ratio of 1.52 to 1.9 is recommended to control the primary mode of
solidification and prevent solidification cracks in type 304L while the Creq/Nieq ratio of 1.42 to
1.9 is recommended for type 316L stainless steel. A lower limit of 50 ppm sulfur is
recommended to avoid possible lack of penetration. The ranges should be validated by welding
Background
A small amount of carbon alloyed with iron makes steel. Iron is allotropic in that
Page 1
Page 2 WSRC-MS-2001-00544
it exists in at least two distinct crystalline forms, primarily dependent upon temperature. At
high temperatures the Face Centered Cubic (FCC) crystal structure of iron is stable and the term
used to describe this phase is austenite. At very high and low temperatures the Body Centered
Cubic (BCC) structure is the more stable phase and is given the name of delta and alpha ferrite,
respectively. Rapid cooling with the attendant high solidification rates from the molten steel
can "quench in" the normally higher temperature phase. Chemical additions (i.e., alloying
additions) can also alter the temperature ranges where these phases are the most stable. Nickel
atoms are nominally the same size as iron atoms and arrange themselves in the FCC structure
over a large temperature range. Therefore, substitution of nickel atoms for iron atoms has the
effect of stabilizing the austenite phase down to low temperatures. Chromium atoms are BCC
and therefore a large substitutional addition of chromium to the steel has the effect of stabilizing
the ferrite phase. Carbon and nitrogen atoms are smaller and occupy interstitial sites between
the primary atoms in a given crystal. The unit cell structure of the austenite phase
accommodates these interstitial atoms more readily than the unit cell structure of the ferrite
phase. Therefore, carbon and nitrogen are very strong austenite stabilizers at relatively small
volume fractions. Sulfur and phosphorus are considered trace impurity elements, remnant from
primary and secondary processing. Steel can hence be considered an amalgamation of chemical
elements within small crystals (grains) along with their accompanying grain boundaries. The
Stainless steels typically contain greater than 12 weight percent chromium for oxidation
and corrosion resistance. Chromium forms a tenacious oxide layer on the surface of the steel,
imparting a passivation layer to provide corrosion resistance. AISI∗ types 304L and 316L (L for
∗
American Iron and Steel Institute
Page 3 WSRC-MS-2001-00544
low carbon content) stainless steels (SS) are iron based austenitic stainless steels with the
manufactured. These alloys have adequate Ni (8 percent minimum) to be fully austenitic in the
as-formed condition with adequate Cr (18 percent minimum) for corrosion resistance. A loss of
free or un-compounded chromium can drastically reduce the local corrosion protection leading
to preferential intergranular attack or heat affected zone (HAZ) attack, with a concomitant
depletion of chromium is due to the formation, growth, and precipitation of chromium carbide
particles in the grain boundaries when and where-ever the steel encounters temperatures in the
range of about 4500 C to around 8500 C, most notably in the HAZ of a weld. By simply reducing
the amount of carbon available for the chemical reaction to occur with chromium, the L grades
of stainless steel result in enhanced weld HAZ resistance to sensitization. The effects of
sensitization are a loss in corrosion resistance due to chromium depletion and a loss of fracture
toughness due to precipitation of complex carbides within and along HAZ grain boundaries.
In general, types 304L and 316L SS are readily weldable with common arc processes
such as; gas tungsten arc welding (GTAW), shielded metal arc welding (SMAW), gas metal arc
welding (GMAW) processes and other techniques1 . These traditional arc processes produce
moderate rates of heat input with correspondingly moderate cooling and solidification rates and
will generally produce a crack free weldment with acceptable microstructure in typical heats of
types 304L and 316L SS. However, a full understanding of specific interactions of primary and
trace elemental constituents of the alloys that can affect the weldability is helpful in ensuring
proper material specification. This article will address some of the chemistry effects as they
relate to welding for ASME SA 240, types 304L and 316L SS. The concepts described in this
Page 4 WSRC-MS-2001-00544
article for ASME SA 240 type 304L, can be applied to other specifications and types of 300
series SS.
When arc welding the 300 series stainless steels, which are fully austenitic in the as-
formed condition1 , research has shown that small amounts of primary delta ferrite retained in
the weld microstructure at room temperature reduce the hot cracking tendencies. Filler
materials are designed by appropriately alloying the iron based material to ensure a minimum
amount of retained ferrite in the completed weldment. For autogenous weldments (no filler
material added) the base metal chemistries of the components to be welded contingent with a
given proportion of dilution will determine, for a particular cooling rate, the composition and
hence the microconstituents of the weld microstructure. The microstructure and thermal history
important for predicting the integrity of the proposed weld from a solidification cracking
perspective. For arc welding processes, the cracking tendency can be predicted by considering
the relative amounts of the ferrite stabilizers to the amount of austenite stabilizers. While for
fusion welding processes with either very high cooling rates (as for the high energy density
beam processes such as laser and electron beam) or very low cooling rates (as with the in-situ
casting processes) the weld nugget morphology becomes more important than the primary mode
of weld pool solidification. The beam processes are best suited to applications that favor very
high aspect ratio weld nuggets (high depth to width ratio). When the beam processes are used
to affect nugget morphologies similar to that of GTAW, then the same concerns over weld hot
cracking and loss of puddle control associated with material constituents similarly apply. With
high energy beam welding (laser or electron beam) performed in the non-traditional low aspect
Page 5 WSRC-MS-2001-00544
ratio mode, loss of puddle control can be more significant than with the arc processes as there
are no arc forces present that can dominate the weld pool stirring mode.
The concept of nickel equivalence (Nieq) is the term used for the cumulative effects of
while chromium equivalence (Creq) is similarly the term used for the ferrite stabilizers. The ratio
of the two terms, i.e., equivalency ratio, is usually taken as Creq/Nieq, based on the actual
chemical composition of the steels involved, and can be used as a quantitative indicator for
predicting the primary mode of solidification for arc (fusion) welded 300 series stainless steel.
Since a small but finite amount of ferrite in the finished weldment is desired, weld pool
cracking during welding. The customary material specifications used in the procurement of the
various types of 304L and 316L allow for a wide range in primary alloy concentration levels.
From a commercial economics perspective the practical chances of obtaining a base material
rich in either chromium or nickel is small. However, with minimill feed stocks being
increasingly high in recycled content and small in relative batch sizes, the potential for receiving
a base material rich enough in alloy content to cause a weld problem, but otherwise meeting
use for the procurement of 304L and 316L allow for the possibility of receiving material overly
rich in austenite stabilizers such that weld solidification as primary austenite could occur and
Chemistry
Base metal chemistry is an especially important consideration for situations where joints
are made autogenously and where recovery from an unacceptable weld is difficult.
Page 6 WSRC-MS-2001-00544
Specifications for typical products used in nuclear material packages and other pressure vessel
applications, for instance, include, ASME SA 1822 , ASME SA 2403 , ASME SA 3124 all for
304L types, ASME SFA 5.95 for 308L type, and ASME SA 240 and ASME SA312 for 316L
types. The composition ranges of the alloys and their respective specification are listed in Table
1. For simplicity, Creq and Nieq will be calculated for ASME SA 240 types 304L and 316L, and
Equivalence
Many researchers6,7,8,9 have examined the composition and post weld microstructure and
have each determined an equation to calculate the nickel and chromium equivalents. Several of
the equations are listed in Table 2 for the calculation of the Creq/Nieq ratio. The range of
composition can be used to predict the possible solidification modes or final microstructure,
depending on the researcher’s methodology, using any of the equations and the corresponding
constitution diagram, e.g., the Shaeffler6 , Delong7 , or Welding Research Council (WRC) 19929
constitution diagrams.
In this article, the Creq and Nieq are calculated corresponding to the equations empirically
derived from the WRC-1992 equation and then plotted on the appropriate constitution diagram
to determine which microconstituent will initially solidify. In general, it has been found that in
the absence of phosphorus and sulfur limit considerations, a minimum of 4 volume percent
ferrite1 (Ferrite Number FN 4) and a maximum of about 21 are required to prevent solidification
cracking. At very low FN, welds solidify as primary austenite and have significant cracking
tendencies, while at very high FN the weld will solidify fully as ferrite and exhibit some of the
same cracking tendencies as with FN less than 410 . In the high FN ranges large amounts of
sulfur or phosphorus may exacerbate hot cracking tendencies. To reiterate, alloys with ferrite
Page 7 WSRC-MS-2001-00544
contents/numbers between 4 and 21 solidify as primary ferrite with austenite (the region
Based on the welding handbook11 the propensity for encountering weld solidification
cracking decreases dramatically at Creq/Nieq ratios slightly less than 1.5 for equivalence
determined from WRC-1992 calculations, as shown in Figure 1. This Creq/Nieq ratio lower limit
is similar but not identical to that produced by Suutala, et al. 12 who used the Hammar and
Svenson8 (H&S) Creq and Nieq equations. The difference between the “critical” Nieq/Creq ratios
for the WRC-1992 and H&S is due to different coefficients attached to the various elements.
This effect can clearly be seen by applying the H&S and WRC-1992 calculations to a typical
heat of Type 304L SS listed in Table 1. The difference between the two calculations is a few
percent.
Suutala et al. report further on the effects of P and S on weld cracking susceptibility as
shown in Figure 2. Sulfur, although a tramp element, has been shown to be beneficial in small
quantities (greater than ~50 ppm) by enhancing weld penetration. This will be discussed in
SA 240 types 304L and 316L and specification ER308L, the appropriate composition ranges
listed in Table 1 were used at the extreme values for Ni and Cr and the Creq and Nieq were
calculated using the formula from the WRC-1992 equivalence equations. The calculated Creq
and Nieq values and Creq/Nieq ratios presented in Table 4 show minimum and maximum values.
The extreme values form the corners of a rectangle when plotted on the WRC 1992 constitution
The plot for ASME SA 240 type 304L shows that approximately one tenth of the area
(possible compositions that meet specification) enclosed by the limit lines will solidify as
austenite, and additional fourth will solidify as primary austenite, one half will solidify as
primary ferrite, while the balance will solidify as ferrite. Similarly for ASME SA 240 type
316L, approximately one third of the possible compositions will solidify as austenite, one fourth
as primary austenite, one third as primary ferrite and the balance as ferrite. The solidification
mode can be altered with the addition of filler metal that is enriched in ferrite stabilizers. Type
308L is one such alloy, it’s equivalence range is plotted in Figure 5, and it exhibits a
solidification mode that will have either primary ferrite or fully austenite. Thus by suitable in
It should be noted that the maximum carbon content permitted by the ASME SA 240 was
used for the calculations. In addition, due to the absence of specification limits for nitrogen, a
typical amount of 0.07% was assumed. It has been shown that a base metal content of 0.02% N
can increase to 0.1% following a fusion weld operation due to nitrogen pick-up from the
atmosphere1 unless the weld is cooled completely to room temperature under a protective gas
shield. Certified Material Test Reports (CMTR) can necessarily include only the measured
nitrogen content of the unwelded base material. Allowance for the prospect of an increase in
nitrogen content for the weld is made by specifying a slightly higher Creq/Nieq in the initial
procurement recommendation. The nitrogen content taken from the CMTR should therefore be
used "as reported" in calculating the Creq/Nieq ratio. However, actual weld sample testing can be
P and S Effects
susceptibility, one must also consider the effects of sulfur and phosphorus on weld penetration
and weld cracking, as indicated by Figure 2. Small amounts of sulfur (0.005 to 0.026 wt%)
have been associated with improved weld penetration, thus it is prudent to specify a lower limit
to better insure weldability. Descriptions of the mechanisms are beyond the scope of this
document suffice to say that very low sulfur stainless steels (less than ~50 ppm) exhibit poor or
intermittent penetration and unstable weld pool control. Heats with excessive sulfur content
may experience heat affected zone (HAZ) cracking or weld centerline cracking, especially in
alloys rich in austenite stabilizers. The effect of phosphorus is primarily one of fusion zone
cracking13 as opposed to weld puddle control or penetration effects. The cracking tendencies of
P and S tend to be combined and are assumed additive as indicated by Satuula in Figure 212 .
The degree of cracking is largely dependent upon the solidification mode so the Creq/Nieq ratio
in conjunction with the P and S content provides some indication of the cracking potential for a
given solidification rate. This is not a simple "rule of thumb" since Brooks et al.14 demonstrated
a shift in solidification mode from primary ferrite to primary austenite for two alloys with
Creq/Nieq ratios of 1.48 and S contents of 0.04% and 0.11%, respectively. Thus it is prudent to
consider the alloy composition, application, post weld inspection, etc., prior to specifying a
addition, the extent of weld restraint present during welding may affect the amount of hot
cracking. For instance, the with no restraint, hot cracks may not form even with primary
austenite solidification while with some restraint hot cracks can be a problem, as shown in
Figure 6. Other welding variables such as shielding gas composition, flux, filler metal, etc., can
Page 10 WSRC-MS-2001-00544
have a pronounced effect on penetration and can be used to alleviate S and P concerns therefore
institutional specifications used for the manufacture and procurement of commercial grades of
stainless steel. For instance, sulfur is added to the free machining grades of austenitic stainless
steels and from a commercial perspective it is not uncommon to encounter heat lots of material
at the high end of the specifications in S. Conversely, minimill feedstocks are often rich in
recycled content and may contain very low residuals of P and S (<.002%, less than ~20 ppm).
This situation is possible since there are no minimum requirements in the ASTM specifications
on trace element concentration levels. A base material with an extremely low sulfur content
may produce a nearly unweldable alloy (i.e., poor penetration and puddle control) yet one that is
Discussion
The review presented here reveals that general material specifications governing alloy
concentration levels cover a wide range of chemistries and are open to maximum limits
concerning primary alloying agents and minimum and maximum limits on residual elements.
The composition ranges permitted by specification ASME SA 240 for types 304L and 316L SS
alone may result in alloys that promote full austenite or primary austenite solidification modes
in autogenous welds of these materials, these solidification modes are known to be crack
sensitive. While the commercial economic realities preclude the likelihood of obtaining alloys
overly rich in either chromium or nickel, the specifications clearly allow for the possibilities.
Certainly incoming steels have tended to be much cleaner (less tramp element content) and
more refined over the last decade. Based on the theoretically possible rich and clean alloys
Page 11 WSRC-MS-2001-00544
allowed by specification ASME SA 240, for example, the predicted weld morphologies suggest
that during fusion welding a potential exists for conditions that promote solidification cracks
due to primary solidification modes other than primary ferrite. Also, a potential exists for lack
of penetration and uncontrollable weld pool conditions due to surface tension variability with
very low sulfur levels. The types 304L and 316L SS material procurement specifications for
fusion welded products should include, in addition to the institutional specifications, high and
low limits on the calculated Creq/Niea ratio and low limits on the residual sulfur content based on
CMTR data, especially for autogenously welded products. The production weld quality
calculations, based on readily available (often required) CMTR data, represents insurance
against future material chemistry induced welding related rejects or field failures.
As an example, the composition of the “typical” ASME SA 240 type 304L SS listed in
Table 1, was used to calculate the Creq and Nie q and these data were plotted in Figure 3. It can
be seen that this heat of material will solidify as primary ferrite with a ferrite number of
approximately 9. Thus, this heat of material should not be prone to solidification cracks due to
austenitic solidification. The second issue that is of concern is the combination of tramp
elements and Creq/Nieq ratio. The Creq/Nieq ratio for this alloy was determined to be 1.81 and
the P+S was 0.05%, Table 3. Comparing the equivalence ratio to Figure 1 indicates that
cracking susceptibility is low, and plotting this on Figure 2 also indicates that this heat of
Specific Recommendation
To help insure that types 304L and 316L SS are relatively crack insensitive, and yet
weldable, particularly when welded autogenously, limits should be placed on the Cre q/Nieq ratios
Page 12 WSRC-MS-2001-00544
and sulfur content. Using the WRC-1992 constitution diagram and Creq and Nieq equations, a
Creq/Nieq ratio range of 1.5 to 1.9 should be used for type 304L SS. The suggested Creq/Nieq
ratio range for type 316L SS is 1.43 to 1.9; the lower limit is decreased due to the higher
propensity for primary austenite solidification with type 316L SS. For both cases, the sulfur
Lower S limits may be used in conjunction with lower Creq/Nieq ratios, however, lack of
penetration or the need to use filler metal may arise. Welding trials to verify that an appropriate
balance of required properties has been achieved may also be incorporated into the procurement
specification.
Acknowledgements:
The authors would like to acknowledge John Brooks of Sandia National Laboratories for
providing micrographs and technical input and the DOE for support under Contract No. DE-
Table 1. Composition ranges for various specifications of types 304L, 308L, and 316L.
App C Mn P S Si Cr Ni Mo N Fe
Specification Type Max. Max. Max. Max. Max. Max.
SA 182 304L 0.035 2.00 0.045 0.030 1.00 18.00- 8.00- N/A N/A Bal.
(forging) 22.00 13.00
SA 240 304L 0.030 2.00 0.045 0.030 0.75 18.00- 8.00- N/A 0.10 Bal.
(plate) 20.00 12.00
SA 312 304L 0.035 2.00 0.040 0.030 0.75 18.00- 8.00- N/A N/A Bal.
(pipe) 20.00 13.00
SA 312 316L 0.035 2.00 0.04 0.03 0.75 16.0- 10.0- 2.00- N/A Bal.
(pipe) 18.0 15.0 3.00
SA 240 316L 0.03 2.00 0.045 0.03 0.75 16.00- 10.00- 2.00- 0.1 Bal.
(plate) 18.00 14.00 3.00
ER308L sfa5.9 308L 0.030 1.0- 0.030 0.030 0.75 19.50- 9.00- 0.75 N/A Bal.
2.5 22.00 11.00 Max.
Typical SA 240 304L 0.024 1.75 0.031 0.022 0.40 18.41 8.75 0.213 0.033 Bal.
Table 2. List of several possible Creq and Nieq that may be used.
Table 3. Calculated equivalence values for the typical SA 240 listed in Table 1 using both
WRC 1992 and Hammar and Svenson equations.
Table 4. Cr and Ni equivalents for the alloys shown in Table 1 using the extremes of the
composition limits as calculated using the WRC-1992 equivalence equation shown below.
Creq/Ni eq
WRC
Specification Creq Nieq
1992
(Min/Max)
SA 240 18.00 9.45 1.20
(304L) 20.00 15.05 2.12
ER308L 20.25 10.45 1.44
Sfa5.9 22.00 14.05 2.11
SA 240 18.00 11.45 1.06
(316L) 21.00 17.05 1.83
WRC 1992: Creq = Cr + Mo + 0.7Nb:
Nieq = Ni + 35 C + 20 N + 0.25 Cu
Page 15 WSRC-MS-2001-00544
References:
Metals Handbook Volume 6, Welding, Brazing, and Soldering, Metals Park, OH, 1993.
2. Anon, ASME 1998 Section II, AMSE SA-182/SA-182M, “Specification for Forged or
Rolled Alloy-Steel Pipe Flanges, Forged Fittings, and Valves and Parts for High-Temperature
3. Anon, ASME 1998 Section II, AMSE SA-240/SA-240M, “Specification for Heat-
Resisting Chromium and Chromium-Nickel Stainless Steel Plate, Sheet, and Strip for Pressure
4. Anon, ASME 1998 Section II, AMSE SA-312/SA-312M, “Specification for Seamless
and Welded Austenitic Stainless Steel Pipes”, NY, pp. 477-490, 1998
5. Anon, ASME 1998 Section II, AMSE SFA-5.9, “Specification for Bare Stainless Steel
7. DeLong, W.T., G. A. Osram, and E. R. Szumachowski, Weld Journal, 35, pp. 521s-528s,
1956.
Weld Metals: A Modification of the WRC-1988 Diagram”, Welding Research Supplement, pp.
11. Anon, Welding Handbook, Eighth Edition, Volume 4., Materials and Applications Part
12. Takalo, T., Suutala, N., and T. Moisio, "Austenitic Solidification Mode in Austenitic
Stainless Steel Welds", Met. Trans. A, Vol. 10A, pp. 1173-1181, Aug. 1979.
13. Li, L. and R.W. Messler, Jr., The Effect of Phosphorous and Sulfur on the Susceptibility
to Weld Hot Cracking in Austenitic Stainless Steels”, Welding Journal, 71, pp. 171s-179s. May,
1999.
Stainless Steel”, Sandia Report Sand2000-8002, Sandia National Laboratories, August 2000.
Page 17 WSRC-MS-2001-00544
List of Tables:
Table 1. Composition ranges for various specifications of types 304L, 308L, and 316L.
Table 2. List of several possible Creq and Nieq that may be used.
Table 3. Calculated equivalence values for the typical SA 240 listed in Table 1 using both
Table 4. Cr and Ni equivalents for the alloys shown in Table 1 using the extremes of the
composition limits as calculated using the WRC-1992 equivalence equation shown below.
Page 18 WSRC-MS-2001-00544
List of Figures
Figure 2. Cracking susceptibility of 300 series stainless steel based on Cr-Ni equivalence from
For Figures 3 through 5: Area A denotes primary solidification as austenite with no further
Area F denotes primary solidification as ferrite with no further transformation (high Creq).
Figure 3. WRC 1992 Constitution diagram for SA 240 type 304L SS composition range Cr and
Ni equivalence plotted.
Figure 4. WRC 1992 Constitution diagram with SA 240 type 316L Cr and Ni equivalence range
indicated.
Figure 5. WRC 1992 Constitution diagram with SFA-5.9 (type 308L) ER308L compositions
ranges indicated.
Figure 6. Micrograph showing solidification cracks in an alloy that solidified with primary
austenite.
Page 19 WSRC-MS-2001-00544
Figure 3. WRC 1992 Constitution diagram for SA 240 type 304L SS composition range Cr
and Ni equivalence plotted.
For Figures 3 through 5: Area A denotes primary solidification as austenite with no further
transformation down to room temperature (high Nieq).
Area F denotes primary solidification as ferrite with no further transformation (high Creq).
Page 22 WSRC-MS-2001-00544
Figure 4. WRC 1992 Constitution diagram with SA 240 type 316L Cr and Ni
equivalence range indicated.
Page 23 WSRC-MS-2001-00544
Figure 5. WRC 1992 Constitution diagram with SFA-5.9 (type 308L) ER308L
compositions ranges indicated.
Page 24 WSRC-MS-2001-00544
Figure 6. Micrograph showing solidification cracks in an alloy that solidified with primary
austenite.