P. S. Korinko and S. H. Malene
Westinghouse Savannah River Company
Aiken, SC 29808

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

The susceptibility of austenitic stainless steels to solidification cracking and lack 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 Cr_eqivalence to Ni_eq 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 Cr_eq/Ni_eq 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 trials for specific weld processes and applications.

A small amount of carbon alloyed with iron makes steel. Iron is allotropic in that 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 finer the grain-size, the stronger and tougher is the steel¹.

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 (American Iron and Steel Institute) types 304L and 316L (L for low carbon content) stainless steels (SS) are iron based austenitic stainless steels with the compositional ranges shown in Table 1, depending on the specification to which it is 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 degradation of mechanical properties. Such a condition is termed "sensitization"¹. The 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 450⁰C to around 850⁰C, 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 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 determine the physical properties and serviceability of the resulting weldment.

The primary mode of solidification of the weld, be it austenite or ferrite, is therefore 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 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 (Ni_eq) is the term used for the cumulative effects of austenite stabilizing elements as a weighted summation of their respective concentration levels while chromium equivalence (Cr_eq) is similarly the term used for the ferrite stabilizers. The ratio of the two terms, i.e., equivalency ratio, is usually taken as Cr_eq/Ni_eq, 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 solidification as primary ferrite is preferred to prevent the likelihood of encountering hot 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 specification, is perceived as a real possibility. Virtually all standard specifications in current 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 lead to hot cracking in a production weld.

Base metal chemistry is an especially important consideration for situations where joints are made autogenously and where recovery from an unacceptable weld is difficult. 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, Cr_eq and Ni_eq will be calculated for ASME SA 240 types 304L and 316L, and for ASME SFA 5.9 type 308L.

Many researchers^6,7,8,9 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 Cr_eq/Ni_eq 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 Cr_eq and Ni_eq 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 contents/numbers between 4 and 21 solidify as primary ferrite with austenite (the region indicated by FA in Figures 3 through 5).

Based on the welding handbook¹¹ the propensity for encountering weld solidification cracking decreases dramatically at Cr_eq/Ni_eq ratios slightly less than 1.5 for equivalence determined from WRC-1992 calculations, as shown in Figure 1. This Cr_eq/Ni_eq ratio lower limit is similar but not identical to that produced by Suutala, et al.¹² who used the Hammar and Svenson⁸ (H&S) Cr_eq and Ni_eq equations. The difference between the "critical" Ni_eq/Cr_eq 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 more detail later.

To evaluate the possible solidification morphologies allowed by the specification ASME 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 Cr_eq and Ni_eq were calculated using the formula from the WRC-1992 equivalence equations. The calculated Cr_eq and Ni_eq values and Cr_eq/Ni_eq 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 diagram as shown in Figures 3 through 5.

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 arc alloying, dilution, the solidification mode can be modified.

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 Cr_eq/Ni_eq in the initial procurement recommendation. The nitrogen content taken from the CMTR should therefore be used "as reported" in calculating the Cr_eq/Ni_eq ratio. However, actual weld sample testing can be considered for critical applications.

In addition to considering the solidification mode of 300 series SS for cracking 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 Cr_eq/Ni_eq 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 Cr_eq/Ni_eq 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 possibly tighter range of allowable compositions than permitted by the specification. In 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 have a pronounced effect on penetration and can be used to alleviate S and P concerns therefore these considerations should be reviewed in the context of specific applications.

Often, serviceability and machinability considerations rather than weldability drive 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 crack insensitive when fusion welded, as suggested by Figure 2.

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 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 Cr_eq/Ni_ea ratio and low limits on the residual sulfur content based on CMTR data, especially for autogenously welded products. The production weld quality assurance enhancement promulgated through the implementation of such chemical content 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 Cr_eq and Ni_eq 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 Cr_eq/Ni_eq ratio. The Cr_eq/Ni_eq 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 material is not crack susceptible.

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 Cr_eq/Ni_eq ratios and sulfur content. Using the WRC-1992 constitution diagram and Cr_eq and Ni_eq equations, a Cr_eq/Ni_eq ratio range of 1.5 to 1.9 should be used for type 304L SS. The suggested Cr_eq/Ni_eq 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 range should be 0.005 to 0.030% for acceptable penetration.

Lower S limits may be used in conjunction with lower Cr_eq/Ni_eq 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.

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-AC09-89SR18035 and DE-AC09-96SR18500.

Table 1. Composition ranges for various specifications of types 304L, 308L, and 316L.

Table 2. List of several possible Cr_eq and Ni_eq 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.

Specification	Creq	Nieq	Creq/Nieq WRC 1992 (Min/Max)
SA 240 (304L)	18.00	9.45	1.20
	20.00	15.05	2.12
ER308L	20.25	10.45	1.44
Sfa5.9	22.00	14.05	2.11
SA 240 (316L)	18.00	11.45	1.06
	21.00	17.05	1.83

1. Brooks, J. A., and J. C. Lippold, "Selection of Wrought Austenitic Stainless Steels", 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 Service", NY, pp. 229-249, 1998.
   
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 Vessels," NY, pp. 363-370.i, 1998.
   
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. ASME 1998 Section II, AMSE SFA-5.9, "Specification for Bare Stainless Steel Welding Electrodes and Rods", NY, pp. 203-227, 1998.
   
6. Schaeffler, A. L. "Constitution Diagram of Stainless Steel Weld Metal", Metal Progress, 56, pp. 680-680B, November 1949.
   
7. DeLong, W.T., G. A. Osram, and E. R. Szumachowski, Weld Journal, 35, pp. 521s-528s, 1956.
   
8. Hammar, O. and U. Svensson, "Influence Of Steel Composition on Segregation and Microstructure During Solidification of Austenitic Stainless Steels", Solidification and Casting of Metals, The Metals Society, London, pp. 401-410, 1979.
   
9. Kotecki, D. J. and T. A. Siewert, "WRC-1992 Constitution Diagram for Stainless Steel Weld Metals: A Modification of the WRC-1988 Diagram", Welding Research Supplement, pp. 171s-177s, May 1992.
   
10. Hull, F.C., Welding Journal, 46, pp. 399s-409s, 1967.
    
11. Anon, Welding Handbook, Eighth Edition, Volume 4., Materials and Applications Part 2., AWS, Miami, FL. 1998.
    
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.
    
14. Brooks, J. A. , S. H. Goods, and C. V. Robino, "Weld Properties of a Free Machining Stainless Steel", Sandia Report Sand2000-8002, Sandia National Laboratories, August 2000.

Figure 6. Micrograph showing solidification
cracks in an alloy that solidified with primary austenite.