
What welding metallurgy and the heat-affected zone (HAZ) are
Every fusion weld creates three regions. The fusion zone, or weld metal, melts and re-solidifies from filler plus diluted parent metal. Next to it sits the heat-affected zone, or HAZ, which never melts but still gets pushed through a steep thermal cycle, hot enough to change its microstructure, chemistry, and corrosion behavior. Past that is unaffected parent metal. The corrosion resistance a buyer specified belongs to the plate chemistry. Welding rewrites that chemistry locally, in the HAZ and the weld pool, which means the joint can corrode or crack long before the surrounding plate ever does.
Why the HAZ decides corrosion performance
A corrosion-resistant alloy resists attack because chromium (and, in the nickel grades, molybdenum and nickel) stays in solid solution, ready to build the passive film. The HAZ thermal cycle can pull chromium out of that solution as carbides. It can leave residual tensile stress. Or it can form a heat-tint oxide that strips chromium from the surface. In every case, the alloy ends up locally depleted. That's why the weld is the weakest link: the governing rule is that the weld and its HAZ must survive the same service the parent metal was specified for. And that starts at material selection, not at the welder's hand. This hub explains what welding does across the stainless and nickel-alloy families we stock, then routes you to the grade-by-grade procedures.
How welding degrades corrosion resistance: the mechanisms
Sensitization and weld decay
Austenitic stainless steel that sits in the 425–815 °C range during a weld thermal cycle lets carbon team up with chromium, forming chromium carbide (Cr₂₃C₆) at the grain boundaries. The metal right next to those boundaries ends up starved of chromium and can't passivate anymore. That's why a thin band in the HAZ starts corroding between the grains, engineers call it weld decay. It's really just a specific case of sensitization / intergranular corrosion, one that shows up in the HAZ specifically.
You fix this at the purchasing stage, not after the fact. Pick a low-carbon "L" grade (C ≤ 0.03%), go with a titanium- or niobium-stabilized grade like 321/347, or plan on a post-weld solution annealing cycle to redissolve the carbides. Filler selection plays into this too, and that side of the problem gets its own treatment in weld decay and filler selection.
Hot (solidification) cracking
Fully austenitic weld metal has no delta ferrite to tie up sulfur and phosphorus. That makes it prone to solidification cracking as the weld pool freezes, since low-melting S/P films open along the dendrite boundaries once shrinkage stress sets in. Standard austenitic stainless welds control this by balancing the weld-metal chemistry to hold roughly 3–10 FN (Ferrite Number) of delta ferrite, a figure predicted from the Creq/Nieq ratio on the WRC-1992 (Schaeffler-type) constitution diagram. Alloys like 904L and Incoloy 825, being fully austenitic, can't carry any residual ferrite. Hot cracking in these cases gets held off a different way: over-alloyed or nickel-base fillers, low heat input, and clean joints.
Chloride stress-corrosion cracking at welds
Austenitic stainless steels are susceptible to chloride stress-corrosion cracking (SCC), and a weld concentrates the two ingredients it needs: residual tensile stress from solidification shrinkage and, if the grade sensitized, a degraded HAZ. The weld line therefore often cracks before the parent plate in hot chloride service. Higher-nickel grades raise the SCC threshold, super-austenitic 904L outperforms 316L here, and stress relief or a full solution anneal removes the residual-stress driver.
Heat tint and oxide scale
The straw-to-blue discolouration beside a stainless weld is a chromium-rich oxide. Forming it pulls chromium out of the metal immediately beneath, leaving a chromium-depleted surface layer. That layer pits and starts crevice attack far more readily than clean passive metal does. Back-purging the root with argon limits the oxide ("sugaring" is the un-purged extreme on the back face). Whatever heat tint remains has to come off through pickling and passivation so the passive film can be restored.
Hydrogen cracking and HAZ hardness (the carbon/PVQ side)
On carbon and pressure-vessel plate such as A516 Gr 70 pressure-vessel plate, the weld risk shifts from chromium chemistry to a hard, hydrogen-charged HAZ. A fast-cooled, high-hardness HAZ, paired with diffusible hydrogen from the arc, produces delayed hydrogen-induced cracking. Preheat and low-hydrogen consumables slow the cooling. Post-weld heat treatment (PWHT) bakes the hydrogen out. For sour (H₂S) service, the NACE MR0175 weld hardness cap applies: typically no more than 22 HRC across weld and HAZ. That's the reason sour-service vessels call for HIC-tested plate and qualified weld procedures.
What controls weld integrity: the influencing factors
A specifying buyer locks in most of these decisions right when the order is placed:
- Carbon content. An "L" grade (C ≤ 0.03%) starves the sensitization reaction outright. A stabilized grade (321/347) takes a different route: it ties carbon up as Ti/Nb carbides instead of letting chromium carbide form.
- Heat input and interpass temperature. Lower heat input and a controlled interpass temperature cut the time spent in the 425–815 °C band and keep distortion down.
- Filler over-matching. The consumable needs to at least match the parent chemistry, and for the nickel grades it should run richer still, so the weld metal is never the first thing to corrode.
- Ferrite balance. Aim for 3–10 FN using the Creq/Nieq ratio on WRC-1992. That's what gives austenitic welds their resistance to hot cracking.
- Joint cleanliness and back-purging. Oil, moisture, an un-purged root, any of these will drive porosity, sugaring, and heat tint.
- Post-weld treatment. Solution annealing, stress relief, or PWHT comes first, then pickling and passivation. Together they restore both the structure and the surface.
Choosing the filler: matching versus over-alloyed
The filler decision follows directly from the parent grade. Standard austenitic stainless steels pair fine with matching fillers. Super-austenitic and nickel-iron-chromium grades are different: they need over-alloyed fillers, because preferential segregation during solidification would otherwise leave the weld metal leaner in molybdenum and nickel than the plate it joins.
| Grade | Family | Matching or over-alloyed? | Recommended filler (AWS class) | Why |
|---|---|---|---|---|
| 304 / 304L | Austenitic SS | Matching | ER308L (ER347 for stabilized) | 308L keeps weld-metal carbon low; retains 3–10 FN ferrite for hot-crack resistance |
| 316 / 316L | Austenitic SS | Matching | ER316L | Carries the 2–3% Mo into the weld so pitting resistance is not lost at the joint |
| 904L | Super-austenitic SS | Over-alloyed | ER385 (904L / 20Cb) or Ni-base (e.g. FM 625) | Compensates for Mo/Ni segregation; fully austenitic, so no ferrite is available |
| Incoloy 825 | Ni-Fe-Cr alloy | Over-alloyed (Ni-base) | ERNiFeCr-1 (FM 65) | Matches the Ni-Fe-Cr chemistry; resists hot cracking in a fully austenitic weld |
| A516 Gr 70 | PVQ carbon steel | Matching, low-hydrogen | Low-H carbon-steel filler (+ preheat / PWHT) | Controls HAZ hardness and diffusible hydrogen for sour-service limits |
Why nickel alloys need over-alloyed or Ni-base fillers
Nickel-rich weld pools solidify fully austenitic, and molybdenum and niobium segregate heavily into the last liquid to freeze. A filler matched to the base metal's composition would leave the dendrite cores short on Mo and Ni compared to the parent plate. That's the region that corrodes first. Over-alloyed fillers like ER385, FM 625, and ERNiFeCr-1/FM 65 push the base chemistry up enough that even the leanest part of the finished weld still outperforms the parent metal on corrosion resistance.
Which grades and families this affects
Weldability shows up in every grade we stock. The table maps each focus grade to its HAZ risk, the fix on the purchasing side, and the grade hub with its how-to-weld procedure.
| Grade | Family | Principal HAZ / weld risk | The fix (set at purchase) | Where to go next |
|---|---|---|---|---|
| 304 / 304L | Austenitic SS | Sensitization / weld decay | Specify 304L (C ≤ 0.03%) | 304L · how to prevent sensitization |
| 316 / 316L | Austenitic SS | Weld decay; chloride SCC at the seam | 316L for welded work; ER316L filler | 316L · how to weld 316 stainless |
| 321 / 347 | Austenitic SS (stabilized) | Knife-line attack if poorly controlled | Ti/Nb stabilization resists HAZ sensitization | austenitic stainless steel |
| 904L | Super-austenitic SS | Hot cracking; Mo/Ni dilution at weld | Over-alloyed ER385 / Ni-base filler | super-austenitic 904L · how to weld 904L |
| Incoloy 825 | Ni-Fe-Cr alloy | Solidification cracking (fully austenitic) | ERNiFeCr-1 (FM 65); Ti stabilization resists sensitization | Incoloy 825 · how to weld Incoloy 825 |
| A516 Gr 70 | PVQ carbon steel | Hard HAZ / hydrogen cracking | Preheat, low-H filler, PWHT; normalized/HIC plate | A516 Gr 70 · specifying A516 (preheat/PWHT) |
Austenitic stainless (304/304L, 316/316L) and stabilized (321/347)
These grades do the heavy lifting among austenitic stainless steel options. If a fabrication sits in the sensitization band, or gets welded without a post-weld anneal, go with the L-grade so carbon stays below 0.03%. Need to handle even higher temperatures? Reach for a 321/347 stabilized grade instead. The role of chromium and carbon explains why: keep chromium in solution, and the HAZ stays passive.
Super-austenitic (904L)
Super-austenitic 904L (UNS N08904, 1.4539) is fully austenitic and copper-bearing, with PREN around 34. Its high nickel content raises the chloride-SCC threshold above 316L, which matters for weld survivability, but it must be joined with an over-alloyed filler because no delta ferrite is available to resist hot cracking.
Ni-Fe-Cr alloys (Incoloy 825)
Incoloy 825 (UNS N08825, 2.4858) is a titanium-stabilized nickel alloy whose stabilization resists HAZ sensitization, but whose fully austenitic weld pool still demands the matching ERNiFeCr-1 (FM 65) consumable. Note the designation: it is Incoloy/Alloy 825, not "Inconel 825", getting that right on the PO and MTC avoids receiving-inspection disputes.
PVQ carbon (A516 Gr 70)
A516 Gr 70 is PVQ carbon steel whose weld behavior is governed by hardness and hydrogen, not chromium chemistry. Low-temperature impact performance is improved by normalizing, and sour service requires the HIC-tested plate (HIC per NACE TM0284) welded under hardness-capped procedures.
How to prevent weld-zone failure
This is the control summary. The step-by-step instructions live on the how-to leaves linked below.
Select a weldable grade right at purchase: a low-carbon "L" or a Ti/Nb-stabilized grade takes the sensitization risk off the table before the first arc is struck.
Match the filler to the parent metal. For 304 and 316, go matching; for 904L and 825, go over-alloyed or nickel-base (see the filler table above).
Restore the structure afterward. A post-weld solution annealing cycle redissolves chromium carbides, while PWHT softens the carbon-steel HAZ.
Clean the surface once that's done. Pickling and passivation strip the heat tint and rebuild the passive film.
Finally, verify. Confirm non-sensitization by ASTM A262, and check weld integrity through PMI, NDT & MTC verification: PMI of weld versus parent, RT/UT, and a hydrostatic test.
How the material we ship supports a sound weld
None of that weldability advice means much if the plate on your dock doesn't actually carry the chemistry it's supposed to. That's why every order ships with an EN 10204 3.1/3.2 mill test certificate: the carbon ceiling on an L-grade, the Ti/Nb addition on 321/347, whatever chemistry makes a grade weldable gets documented and traced, not just assumed. Every plate is genuine mill stock from nameable mills (Aperam, Outokumpu, Jindal, POSCO, Nippon Steel, JFE), and third-party inspection through TÜV SÜD, Bureau Veritas, DNV, SGS, or ABS can be arranged against the order. The low-carbon and stabilized grades recommended here for welded work, namely 304L, 316L, 321, and 347, sit ex-stock and ready for immediate dispatch. So "pick a weldable grade" isn't just advice here; it points to material we can actually ship.
Worked examples
A sensitized 304 bracket. A fabricated 304 support, held near 600 °C in service, developed rust lines running parallel to its welds. Classic weld decay was the cause: chromium-carbide precipitation in the HAZ. At re-order, the fix was to re-specify the part in 304L (C ≤ 0.03%), which eliminates sensitization without needing a post-weld anneal. The common error? Defaulting to plain 304 for any part that will be welded and then held warm.
A cracked fillet weld on Incoloy 825 turned up centerline cracking, and the cause traced back to a near-matching stainless filler that had no business being there. 825 segregates heavily in the weld pool and has no ferrite to fall back on. Run it with the wrong consumable and solidification cracking is pretty much guaranteed. The fix is to call out the filler class right alongside the plate, and to use the correct "Incoloy 825" designation on the PO. Do that, and the rejection doesn't happen. The right filler here is ERNiFeCr-1 (FM 65).
An A516 vessel seam. On a thick A516 Gr 70 shell seam meant for low-temperature duty, the under-specified risk was a hard HAZ and hydrogen cracking. The fix: normalized plate with Charpy impact testing at the design temperature, documented preheat and PWHT, and for sour duty, the HIC-tested variant welded to a hardness cap. Plain as-rolled Gr 70 wouldn't cut it.
Grade-by-grade welding procedures
This hub stays declarative. The step-by-step welding procedures themselves live on the guides.
- how to weld 316 stainless
- how to weld 904L
- how to weld Incoloy 825
- how to prevent sensitization
- specifying A516 (preheat/PWHT)
Welded corrosion-resistant assemblies like these show up in hygienic and pharmaceutical fabrication, pulp and paper, and fertilizer and acid service.
