Austenitic stainless steel looks straightforward to weld — no preheat, no post-weld heat treatment in most structural applications, and readily available filler metal. But the metallurgy hides a specific cracking mechanism that standard WPS documentation often misses: hot cracking in the weld metal caused by insufficient ferrite content. Understanding Ferrite Number and how AWS D1.6 addresses it is not optional for shops welding structural stainless.
What Is Ferrite Number and Why Does It Matter?
Ferrite Number (FN) is a dimensionless quantity that represents the magnetic ferrite content in austenitic stainless steel weld metal. It is measured magnetically using a calibrated instrument on the weld surface or predicted from weld metal composition using diagrams such as the WRC-1992.
The significance: austenitic stainless steel welds solidify in one of two modes depending on chemistry — austenitic or ferritic-austenitic. Welds that solidify in the austenitic mode (very low FN, approaching 0) are highly susceptible to solidification hot cracking. Liquid films form at grain boundaries during solidification; when the weld is under restraint, those films tear rather than fuse, producing cracks that run through the center of weld beads.
Welds with at least 3–5 FN solidify partially as delta ferrite first, which disrupts the continuous grain boundary liquid film network and dramatically reduces hot cracking susceptibility. This is why minimum FN requirements exist — not as a material property goal for the finished joint, but as an insurance mechanism against solidification cracking.
Above approximately 10–15 FN (varying by alloy), ferrite content itself begins to reduce corrosion resistance (ferrite is anodic relative to austenite in many service environments) and can cause sigma phase embrittlement over time at elevated temperatures. The practical target is a narrow window: enough ferrite to prevent cracking, not so much that service properties suffer.
AWS D1.6 Ferrite Control Requirements
AWS D1.6:2017 Structural Welding Code — Stainless Steel governs welding of structural stainless steel components. Unlike AWS D1.1, which focuses entirely on carbon and low-alloy steel, D1.6 must address the unique solidification and corrosion behavior of austenitic alloys.
AWS D1.6 addresses ferrite control primarily through filler metal selection requirements. The code requires that:
-
Filler metal selection must account for dilution effects. The weld metal FN is not simply the FN of the filler rod or wire at face value — it is a function of base metal dilution. Higher dilution (deeper joint, hotter process settings) shifts the weld metal composition toward the base metal chemistry, which may raise or lower the resulting FN depending on the alloy pair.
-
Certified test reports (CTRs) must document weld metal composition. The electrode or wire manufacturer's certification should include actual weld metal chemistry (not just electrode chemistry), from which FN can be predicted using the WRC-1992 diagram if the shop requires verification before welding.
-
Prequalified filler metals for structural grades. D1.6 Table 3.1 lists prequalified filler metal classifications for common austenitic base metals (304/304L, 316/316L, 321, 347, and duplex grades). For joints within those combinations, using a listed filler metal with documented conformance satisfies the filler selection requirement without additional FN testing on every lot.
See AWS D1.6 stainless structural welding procedure overview for the broader qualification framework and the full list of essential variables under D1.6.
The WRC-1992 Diagram in Practice
Before a PQR is run, the WPS writer can predict whether the selected filler metal will achieve adequate FN by using the WRC-1992 (Welding Research Council) diagram — the current industry standard for FN prediction. The diagram maps Cr-equivalent (Cr_eq) and Ni-equivalent (Ni_eq) to predicted FN:
Cr_eq = %Cr + %Mo + 0.7×%Nb
Ni_eq = %Ni + 35×%C + 20×%N + 0.25×%Cu
These values come from the electrode or wire certified test report (filler metal composition) and, when accounting for dilution, a weighted blend of filler metal and base metal chemistries.
For a typical 316L base metal welded with ER316L filler at 25–30% dilution, the resulting weld metal Cr_eq and Ni_eq can be plotted on the WRC-1992 diagram to confirm that FN falls in the 3–10 range before the first arc is struck. If the plot falls in the hot-cracking region (FN < 3), the filler selection must be reconsidered — either a higher-FN wire designation or reduced dilution parameters.
This is a pre-qualification screening step, not a guarantee. The actual PQR coupon weld metal should still be measured with a ferritescope on the deposited weld for confirmation.
What the WPS Must Document
A well-prepared AWS D1.6 WPS for austenitic stainless welding must explicitly address ferrite:
Filler metal specification and classification. The AWS A5.9 (bare wire/rod), A5.22 (flux-cored), or A5.4 (covered electrodes) classification must appear on the WPS. The FN prediction is only as good as the filler metal traceability.
Weld metal FN range or minimum. State the required FN on the WPS: Weld metal FN: 3 minimum (WRC-1992 predicted from CTR chemistry; verify by ferritescope measurement on PQR coupon). This makes the requirement visible to QC and the inspector without digging through the PQR.
Dilution control parameters. Travel speed range, voltage range, and amperage range all affect dilution. Documenting these as ranges (not single values) controls the dilution corridor and keeps the resulting FN prediction valid across the parameter range.
Heat input limits. Excessive heat input on stainless not only reduces corrosion resistance in heat-affected zones (sensitization risk for non-stabilized grades) but also affects solidification mode. The WPS should state a maximum heat input per the joint's service requirements.